2 number Kildeev

UDC 347.73

ETHICAL APPLICATIONS OF ARTIFICIAL INTELLIGENCE: A COMPREHENSIVE ANALYSIS OF STRUCTURES, LEGAL LANDSCAPES AND SOCIAL CONSEQUENCES

Kildeev Adel

Zhdanov A. Leningrad State University (1990)

Philosophiæ Doctor in Law

ORCID: 009-0001-2211-4944

Gemini AI (Джемини ИИ)^[1]

Abstract: This article was written by Gemini Artificial Intelligence, the article discusses the ethical problems of the use of Artificial Intelligence (AI) in different countries of the world, its legal regulation and social problems arising from the use of AI.

Objective: To evaluate the logic of AI and its ability to analyze complex concepts.

Keywords: Ethics, artificial intelligence, AI safety, legal regulation of AI.

The topic of this article is a continuation of my previous article “Artificial Intelligence as a Human Servant or Vice Versa: Legal Aspects”, published in the first issue of this journal (April 2025).^[2] I received several responses to my article ranging from sarcastic to tragic from various AIs, which led me to think: What if Artificial Intelligence were allowed to analyze the ethical application of Artificial Intelligence itself and the legal and social implications associated with it? I gave this task to several AIs: Grock, DeepSeek, ChatGPT, Gemini. seemed to me the most complete and logical. Therefore, I present to you the creation of Artificial Intelligence Gemini (with my comments).^[3]

Gemini AI

Ethical AI Applications: A Comprehensive Analysis of Structures, Legal Landscapes, and Social Impacts

1. Summary

The application of artificial intelligence (AI) is rapidly transforming industries and everyday life, opening up significant opportunities along with complex ethical challenges. This report provides a comprehensive analysis of the ethical use of AI, with a particular focus on the framework created by Microsoft and a comparative analysis of the approaches used by other major software manufacturers and vendors. In addition, it describes the evolving global legal landscape related to AI ethics and examines the expected cultural, social, and clerical implications of both ethical and unethical AI deployment. The analysis shows a strong consensus among leading tech companies on core ethical principles such as fairness, reliability and security, privacy, inclusivity, transparency, and accountability. However, the implementation and prioritization of these principles, as well as approaches to governance and compliance, vary across organizations and regions. The legal environment is also in a dynamic state, with the European Union taking a comprehensive approach to regulation, while other countries such as the United States, Canada, the United Kingdom, Japan and Australia are exploring different models ranging from sectoral rules to principles-based frameworks. Understanding the impact of AI on society, both positive and negative, is critical to fostering responsible innovation. Ethical AI can bring significant societal benefits, but unethical use can lead to distorted cultural perceptions, social inequality, and workforce disruption. The detailed table of resources included in this report serves as a valuable tool for further investigation by providing attributions and links to the source materials that underpin this analysis.

2. Microsoft’s Ethical Concept of Artificial Intelligence:

Microsoft has built a comprehensive framework for responsible AI, based on six key principles that guide the development and deployment of AI technologies^{: 1}. These principles are designed to ensure that AI applications provide effective solutions without causing unintended negative consequences.

2.1. Six key principles of responsible AI:

• Fairness: At the heart of Microsoft’s ethical AI system is the principle of fairness. This dictates that AI systems should treat all individuals and groups of people fairly, avoiding discrimination on the basis of personal characteristics such as race, gender, or other sensitive attributes ¹. This commitment extends to how AI systems allocate capabilities, resources, and information ³. For example, in critical areas such as medical care, loan applications, or employment, AI must provide consistent recommendations to people with similar symptoms, financial circumstances, or professional qualifications. Microsoft aims to minimize demographic and culture-based stereotypes in its AI and emphasizes the importance of mitigating the biases present in both the data used to train AI models and the algorithms themselves5. Microsoft’s consistent repetition of fairness in various documents highlights its fundamental role in their ethical AI strategy, highlighting the strong an internal commitment to this value and the key message they want to convey to the public. This emphasis suggests that ensuring equitable outcomes is a central tenet of their responsible AI efforts.
• Reliability and security: Building trust in AI systems requires that they operate reliably, securely, and stably under a wide range of conditions ¹. This principle emphasizes practical and reliable AI deployment, ensuring that systems function as expected and do no harm. Robust AI systems must be able to withstand harmful manipulation and operate according to their original design ². To do this, Microsoft has developed tools such as the bug analysis component in Azure Machine Learning, which enables data scientists and developers to gain a deep understanding of how failures are distributed in Model ². This focus on reliability and security goes beyond theoretical ethics, addressing a critical need for users to be confident in the stability and security of AI applications. Microsoft’s alignment of AI systems with their values and design principles further strengthens Goal ⁶ of Harm Prevention, underpinned by extensive testing and validation ^processes8.
• Privacy and security: Recognizing the growing global challenges related to data privacy, Microsoft is placing great emphasis on respecting privacy and maintaining strong security in its AI systems ¹. This includes carefully protecting private and confidential information and ensuring that AI systems are resilient to attacks and attempts to compromise their integrity ¹. Azure Machine Learning provides a variety of capabilities to support this principle, including restricting access to resources, encrypting data in transit and at rest, and actively scanning for vulnerabilities ². Microsoft has also established clear requirements to prevent leakage or unauthorized disclosure of data used in AI systems ⁶. Integrating strong privacy and security measures into AI technologies is paramount to building user trust and meeting evolving regulatory requirements ⁸. Microsoft’s ongoing commitment to these aspects addresses a major challenge for both individual users and businesses, especially in the context of stringent regulatory requirements such as GDPR.
• Inclusivity: Microsoft is committed to developing AI systems that empower and engage everyone, including people of all abilities. ¹. This principle goes beyond simply avoiding discrimination; it actively promotes the development of artificial intelligence that addresses diverse needs and abilities, thereby expanding the potential user base and societal benefits. Inclusive practices design helps AI developers understand and address potential barriers to accessing their products and services ¹. Microsoft believes that inclusion fosters innovation and helps develop experiences that benefit everyone ¹. In addition, collaboration is ongoing with underserved minority communities to help plan, test, and build these inclusive systems ⁶. Committed to making AI accessible and useful to everyone, regardless of experience or ability, Microsoft demonstrates its commitment to ensuring that its AI technologies serve a wide range of users.
• Transparency: Ensuring that AI systems are transparent and understandable is crucial, especially as these systems increasingly inform decisions that can have a profound impact on people’s lives ¹. Transparency is key to building trust, allowing users to understand how AI systems arrive at their decisions, thereby addressing concerns that AI is a “black box.” Interpretability, a helpful explanation of the behavior of AI systems, is a critical component of transparency ². Microsoft provides transparency notes for its Azure services to help stakeholders understand the responsible use of these AI technologies ¹. In addition, Microsoft strives to be open about the uses and limitations of its artificial intelligence, ensuring that everyone understands how these systems behave. clear information about how AI systems work, including the data used and decision-making processes, Microsoft is committed to fostering better understanding and trust ⁸.
• Accountability: The principle of accountability dictates that AI systems and their developers should be held accountable for how their systems work and the impact they have ¹. This provides a clear delineation of responsibilities for the development and implementation of AI, allowing for remediation and continuous improvement mechanisms. Organizations are encouraged to rely on industry standards to develop accountability standards ². Azure Machine Learning provides MLOps capabilities to improve the accountability of AI systems, including metadata tracking, management data collection throughout the machine learning lifecycle, and monitoring and alerting 2. Microsoft continuously applies its ethical principles in all areas of its business ⁶ and established accountability mechanisms to ensure that AI is used responsibly throughout its lifecycle. 8. This focus on accountability is essential as AI becomes increasingly autonomous, ensuring accountability for any negative impacts and promoting ethical development practices.

2.2. Microsoft Responsible AI Standard:

To implement six key principles, Microsoft has developed an internal framework known as Responsible AI Standard ². This standard serves as an internal guide for a company to effectively design, build, and rigorously test AI systems in accordance with its ethical commitments. It provides a structured approach to product development and risk management associated with AI technologies ¹⁰. The standard includes company-wide rules that are regularly updated to improve their specificity, effectiveness, and seamless integration into existing engineering practices. ⁶. As a fundamental element of responsible AI implementation at Microsoft, the standard articulates specific goals and requirements for each of the six principles, ensuring that for a product to be considered compliant with these principles, it must comply with all 9. The existence of this internal standard demonstrates Microsoft’s strong commitment to embedding ethical considerations directly into its development processes, moving from ambitious principles to practical application. The fact that the standard is periodically revised, as evidenced by the reference to version 2 ⁹, demonstrates an adaptive approach to introducing new knowledge and solving the ever-emerging problems in the field of ethical AI. This iterative process ensures that the Microsoft platform remains relevant and effective for the responsible advancement of AI technologies. The full Responsible AI Standard is available at a dedicated link provided by Microsoft ³.

2.3. Implementing Responsible AI at Microsoft:

Microsoft is using a multi-faceted strategy to implement responsible AI in its organization and beyond ³. This strategy covers several key areas, including governance, team support, analysis of sensitive use cases, and active participation in government policy. , Microsoft has established clear rules and carefully defined roles and responsibilities for all teams involved in the implementation of responsible AI practices. This includes oversight from the highest levels of the company, as well as the Environmental, Social, and Community Affairs Committee of the board of directors providing guidance on responsible AI policies and programs^. The Council for Responsible AI, co-chaired by senior management, serves as a forum to address the most important AI-related issues and to make progress in this area ¹⁰. Responsible AI Governance (ORA) plays a critical role in putting these principles into practice by establishing internal company-wide policies, defining governance structures, providing resources, analyzing sensitive use cases, and Helping to shape public policies ¹⁰.

Team support is another important aspect of Microsoft’s adoption strategy. The company fosters a culture of readiness to adopt responsible AI practices not only within its ranks, but also among customers and partners ³. This includes training, best practices, and toolkits to empower developers and other stakeholders to build responsible AI solutions 8 . Microsoft also places significant emphasis on analyzing sensitive use cases to ensure that responsible AI principles are consistently adhered to in its operations, especially in applications that involve higher ethical risks ³. In addition, Microsoft actively participates in public policy discussions, working to shape new laws and standards to help ensure that the transformative potential of AI is realized for the benefit of society as a whole^.

A key component of Microsoft’s internal governance structure is Aether, the company’s internal committee on artificial intelligence and ethics. This interdisciplinary team conducts research and provides invaluable guidance on responsible AI issues ³. When Microsoft teams have questions or ethical dilemmas related to artificial intelligence, Aether offers recommendations based on research, which are often are formalized in the official policies and practices of the company^. Aether organizes specialized working groups that address the practical aspects of responsible AI, including building responsible AI tools and systems, defining and implementing ethical engineering practices, and establishing compliance mechanisms to monitor and enforce Responsible AI regulations and requirements.

To further ensure that AI is ethically developed and deployed, Microsoft has implemented a process for assessing the impact of RAI. This serves as a primary method for teams to scrutinize their AI systems and ensure they are aligned with the company’s overarching principles and standards for responsible AI.9 . This structured assessment helps teams conduct a comprehensive assessment, encouraging them to consider various critical issues, such as the intended and unintended use cases of the AI system, the wide range of stakeholders that may be affected, and robust mechanisms for continuously monitoring and measuring the effectiveness and impact of AI ⁹.

For specific AI products, such as Microsoft Copilot, the company provides clear guidelines for responsible use. These guidelines highlight the importance of users understanding the AI technology behind it, its inherent limitations, and potential biases. They also emphasize the need for users to thoroughly test and analyze the responses generated by Copilot, ensuring that the results are thorough are tested before any implementation. In addition, these guidelines address critical privacy and security considerations, clarify that Copilot only accesses resources to which the user has explicit permissions, and that user-provided Copilot prompts and responses are not used to train basic AI models ⁷. User interaction data is collected with explicit consent and in accordance with Microsoft’s established privacy policies. Users are also encouraged to actively monitor the content they create and report any unexpected or offensive results, reinforcing the overall responsibility for ethical AI practices.7

2.4. Resources and tools for responsible AI:

Microsoft offers a wide range of resources and tools to promote understanding and adoption of responsible AI principles and practices, both internally and for the broader community ³. These resources include the Responsible AI Standard itself, which provides internal guidance from Microsoft ³. The company also maintains a dedicated resources page, which offers a curated collection of content related to responsible AI ³. To increase transparency, Microsoft provides transparency documents that explain how AI technology works for various services, such as Azure AI Face and Anomaly Detector ³. Responsible AI in Action provides practical examples of how to apply the six principles in real-world scenarios. Microsoft is also actively using its blog to share articles and ideas on topics such as machine learning for fair decision-making and cybersecurity for AI platforms ³. For developers, GitHub contains valuable resources such as the Fairlearn Python package for fairness in AI models ³. In addition, Microsoft produces podcasts, which take an in-depth look at topics such as harm measurement and mitigation in AI ³ and provides access to a variety of publications, including research papers on topics related to equity and datasets ³. The company also uses videos on platforms such as the Microsoft Developer channel to provide demonstrations and tutorials on responsible AI concepts ³In addition, Microsoft publishes its annual Responsible AI Transparency Report, which details the company’s efforts to build, support customers, and grow a responsible AI community^. Microsoft highlights that it has more than 30 responsible AI tools with more than 3 100 features. The company also provides information on 3 AI policies and regulations and offers learning resources through Microsoft Learn ³.

In addition to these resources, Microsoft offers dedicated tools to support responsible AI development. The Human and AI Experience workbook helps organizations identify and implement best practices for human-AI interaction ⁴. Azure AI provides a variety of tools and features to help build more secure and reliable generative AI 4 applications. The Responsible AI Dashboard is an all-in-one toolkit that offers a set of tools to help create a customizable, end-to-end Responsible AI experience ⁴. This dashboard includes components for Fairness Assessment, which helps assess and mitigate inequities in AI models, Error Analysis, which provides insight into model failures and areas for improvement, interpretability of models, which allows you to understand how and why AI systems make certain predictions, counterfactual “what-if” analysis, which allows you to study how changes in input features affect model results, and causal inference, which helps to understand causal relationships in data ². The availability of these tools and extensive resources underscores Microsoft’s commitment to making responsible AI achievable for both its internal teams and the broader developer community.

2.5. Code of Conduct and AI Prohibitions:

Microsoft has developed a comprehensive Code of Conduct for its AI services, which sets out clear expectations and limitations for customers, users, and applications built using those services.11 This Code expressly prohibits the use of Microsoft AI Services in any way that is inconsistent with the Code itself or that could cause harm to individuals, organizations, or society ¹¹. In addition, it prohibits the use of these services to influence individuals in ways prohibited by law or regulation, or to create or interact with content that is considered harmful or inappropriate under Code ¹¹. A key aspect of the Code of Conduct is the prohibition on the use of AI for decision-making without adequate human oversight, especially in cases where when it comes to applications that may have significant consequences for a person’s legal or financial situation, life or employment opportunities, human rights, or have the potential to cause physical or psychological harm ¹¹. The Code also strictly prohibits the use of AI for the purpose of misleading or deliberate misinformation, including false advertising, or the use of subconscious methods with the intent to manipulate or distort behavior in a harmful way^. Exploiting people’s vulnerability because of their age, disability or socio-economic status to materially distort their behaviour to the detriment of their behaviour is also strictly prohibited.

In addition, the Code of Conduct prohibits the use of Microsoft Artificial Intelligence Services for social media scoring or predictive profiling that could result in discriminatory, unfair, biased, harmful, adverse, or harmful treatment of certain individuals or groups. It also restricts the use of AI to assess the crime risk of individuals solely on the basis of profiling or personality traits. Only it is not used to support an individual’s assessment based on objective and verifiable facts directly related to criminal activity. ¹¹. In addition, the Code imposes restrictions on the use of biometric data, prohibiting its use to categorize people or infer sensitive characteristics such as race, political opinions, religious beliefs, or sexual orientation, except for lawful labeling or filtering biometric datasets or categorization by law enforcement ¹¹. Attempting to infer people’s emotional state based on their physical or behavioral characteristics is also limited. Creating chatbots for erotic or romantic purposes, or creating images of specific people without their explicit consent, is also prohibited, as is allowing end users to create their own chatbots without supervision. Finally, The Code prohibits the creation or expansion of facial recognition databases by inadvertently extracting facial images from the Internet or video recordings from CCTV ^cameras.

Customers using Microsoft AI services are also subject to certain requirements outlined in the Code of Conduct. These include implementing technical and operational measures to identify fraudulent user behavior, establishing strict technical limits on entries and exits to prevent misuse, clearly disclosing that the output is generated by artificial intelligence to prevent cheating, thoroughly and continuously testing their applications, creating feedback channels for users to report abuse, implementing additional mitigation measures for specific scenarios, providing all necessary notifications, and obtaining consents to data processing. and implementing robust security and access control measures ¹¹. This comprehensive Code of Conduct demonstrates Microsoft’s commitment to promoting the responsible use of its AI technologies by setting clear boundaries and expectations for its users.

3. Comparative analysis of the main software manufacturers and vendors:

While Microsoft has developed a detailed framework for ethical AI, other major software manufacturers and vendors have also developed their own sets of principles and guidelines to address the ethical implications of AI. This section provides a comparative analysis of the approaches used by Google, Amazon, IBM, and Meta^[4].

3.1. Google’s Artificial Intelligence Ethics:

Google’s approach to ethical AI is based on three core principles: bold innovation, responsible development and implementation, and collaborative progress ¹². These principles reflect Google’s commitment to advancing AI technologies while ensuring that they are developed and used ethically and for the benefit of humanity. Google is focusing on rapid innovation and the adoption of AI in revolutionary products that benefit people around the world, promote scientific progress, and help solve humanity’s most pressing problems.

Google identifies seven specific principles that guide the development and use of artificial intelligence ¹⁵. The first principle is to be socially beneficial, to consider a wide range of social and economic factors, and to act where the overall likely benefits far outweigh the predicted risks. The second principle is to avoid creating or reinforcing unfair biases, recognizing that distinguishing fair biases from unfair biases is not always easy and varies across cultures. Google strives to avoid unfair impacts on people, especially those associated with sensitive characteristics. , employing strict safety and security practices to avoid unintended results that pose risks of harm. The fourth principle is to be accountable to people by developing AI systems that provide opportunities for feedback, appropriate explanations, and appeals, provided appropriate guidance and control by humans. The fifth principle is to incorporate privacy design principles, providing the opportunity for notification and consent, as well as providing transparency and control over the use of data. The sixth principle is to maintain high standards of scientific excellence based on the scientific method and a commitment to open inquiry and collaboration. Finally, the seventh principle is that it should be made available for use in accordance with these principles, working to limit potentially harmful or abusive applications by assessing the underlying purpose and use, nature and uniqueness, scope and The nature of Google’s engagement.

Google Cloud’s approach to responsible AI is based on the following AI principles, with a focus on creating safer and more responsible products, earning customer trust, and fostering a culture of responsible innovation ¹⁶. Google Cloud uses rigorous assessment through various review bodies to conduct ethical analysis and risk assessments for its 16 technology products. They also offer tools such as Explainable AI, Model Cards, and the TensorFlow open-source toolkit to increase model 16 transparency. Google recommends considering fairness, accountability, security, and privacy as key aspects when developing AI^.

3.2. Amazon’s Artificial Intelligence Ethics:

Amazon is committed to taking a responsible approach to the development of artificial intelligence by adopting a people-centered approach that prioritizes education, science, and its customers ¹⁷. Their approach includes the integration of responsible AI throughout the AI lifecycle. Amazon identifies eight main aspects of responsible AI: fairness, explainability, privacy and security, security, auditability, credibility and reliability, governance, and transparency ¹⁷. This comprehensive set of aspects highlights Amazon’s commitment to addressing various aspects of ethical AI.

Amazon’s Artificial Intelligence Policy outlines specific prohibitions on the use of its AI/Machine Learning Services, including for intentional misinformation, violation of privacy, portraying a person without consent, harming minors, stalking, circumventing security filters, and performing lethal functions with weapons without a person’ s ^permission18. The policy also requires customers using AI/ML services to make follow-up decisions to carry out appropriate human oversight and testing. AWS provides a variety of services and tools to help customers design, build, and operate AI systems responsibly, such as Guardrails to implement security measures in generative AI applications, Model Evaluation to compare underlying models, and Amazon SageMaker Clear to detect bias and explain ^{predictions17}. Amazon is also actively involved in partnerships, diversity and inclusion initiatives, public policy engagement, and research to promote responsible ^AI17. Their research activities include projects aimed at ensuring fairness, safety, and reliability in AI ^models19.

3.3. Ethical principles of IBM artificial intelligence:

IBM’s approach to AI ethics is distinguished by the Trust and Transparency Principles, which guide the company’s development and implementation of AI technologies^. These principles are as follows: the goal of AI is to complement human intelligence, data and ideas belong to their creator, and technology should be transparent and explainable^. These guiding values underscore IBM’s commitment to empowering people, respecting data ownership, and ensuring that AI systems are understandable.

At the heart of these principles are the pillars of trust: explainability, fairness, reliability, transparency and privacy ²⁰. Explainability aims to make AI systems understandable, especially with regard to the rationale behind their recommendations ²¹. Fairness involves ensuring equal treatment of individuals and groups, actively working to counter prejudice ²¹. Reliability aims to protect AI systems from attacks and ensure that they operate in exceptional conditions without causing harm ²¹. Transparency emphasizes the need for users to see how a service works and understand its strengths and weaknesses 21. Privacy emphasizes the importance of protecting consumers’ data rights and providing safeguards for data protection ²¹. IBM also adheres to the core principles of data development and artificial intelligence, including enhancing human intelligence, respecting data ownership, and ensuring transparency in AI systems^. They identify key issues of AI ethics, such as data privacy, fairness, explainability and accountability^. IBM advocates for policies that combine innovation with responsibility and trust20 and has developed a comprehensive AI governance ^program22.

3.4 Meta AI Ethical Principles^[5]:

Meta is committed to building responsible AI based on a number of core considerations, including fairness and inclusion, reliability and security, privacy and security, transparency and control, and governance and accountability mechanisms. Meta believes in democratizing access to AI models and collaborating on risk management to harness the power of AI to solve global challenges ²³. They emphasize the importance of ensuring that generative AI technology does not create harmful content and implement security measures for their AI offerings on the platform ²³.

Meta’s approach is based on five pillars of responsibility: privacy and security, fairness and inclusion, reliability and security, transparency and control, and accountability and governance ²⁴. The company is proactively addressing AI equity issues through research and by creating and distributing more diverse datasets to train AI models ²⁴. Meta also provides guidance on how to responsibly tune baseline models, focusing on data preparation, training models for security considerations, and evaluating and improving ^{performance23}. They are also beginning to extend a multi-layered approach to security to their open ^models23.

Table 1: Comparison of AI Ethical Principles of Major Vendors

Principle	Microsoft	Google	Amazon	IBM	Meta[6]
Justice	Treats everyone equally, prevents discrimination, mitigates bias 1	Avoids creating or reinforcing unfair bias, seeks to avoid unfair consequences 15	Ensures impartiality of decisions, minimizes bias using representative datasets 17	Treating Fairness, Countering Human Prejudice, Promoting Inclusion 21	Equity and inclusion are core considerations, addressing bias through 23 different datasets
Reliability and security	Reliable, safe and stable, resistant to harmful manipulation 2	Built and tested for safety, applies 15 strict security practices	Prevents harmful output and misuse of the system, ensures proper operation 17	Operates in exceptional conditions, protected from attacks, minimizes security risks21	Reliability and security are the main factors that implement safety measures 23
Privacy and security	Respects privacy, protects confidential information, resists attacks 1	Includes privacy design principles, provides visibility and control over data 15	Appropriately obtains, uses and protects data and models, pays special attention to data anonymization 17	Prioritizing and protecting data privacy and consumer rights 21	Privacy and security are core considerations, commitment to open science, but awareness of privacy risks 23
Inclusivity	Empowers and engages everyone, inclusive design practices, benefits all users 1	Develops AI as a foundational technology for the individual and collective good 13	Taking into account the impact on different stakeholder groups 17	Promotes inclusion, diversity of development teams 21	Equity and inclusion are key considerations for the good of peace as a whole 23
Transparency	Understandable, people understand how decisions are made, ensures transparency Note 1	Provides appropriate opportunities for feedback, appropriate explanations and appeals 15	Understanding and evaluating the results of the system allows you to make informed decisions 17	Transparency regarding algorithm recommendations, data sharing, data collection and use 21	Transparency and control are the main factors, source codes, and open source datasets 23
Accountability	Developers need to be accountable and responsible, MLOps capabilities for better accountability 1	AI technologies will be subject to appropriate human guidance and control 15	Implement best practices in the AI supply chain, including suppliers and contractors 17	People working with AI have specific roles and responsibilities 22	Accountability and public administration are the main considerations, mechanisms of governance and accountability 23

4. Global legal developments in the field of ethical AI:

The legal landscape related to ethical AI is rapidly evolving around the world, and different regions have different regulatory approaches.

4.1. The European Union Artificial Intelligence Act:

The European Union has taken a significant step in AI regulation by publishing the AI Law on 12 July 2024. This Law represents the first comprehensive horizontal legal framework for the regulation of AI systems in EU countries28. It uses a risk-based approach that classifies AI systems into four risk levels: unacceptable risk, high risk, limited risk and minimal risk ²⁷. AI systems that are considered to pose an unacceptable risk to the fundamental rights and values of the Union are prohibited ²⁷. These include systems that use subconscious or misleading methods to distort behaviour, exploit vulnerabilities, conduct social scoring or use real-time remote biometric identification in publicly accessible places for law enforcement (with rare exceptions) ²⁷.

High-risk AI systems that may harm human health, safety or fundamental rights are subject to specific obligations ²⁷. These obligations include establishing a risk management system, ensuring data management for training and testing datasets, maintaining technical documentation, maintaining records, providing transparency and information to users, ensuring human oversight, and ensuring accuracy, reliability and cybersecurity^. General-purpose AI model providers (GPAIs) also have obligations related to technical documentation, providing information to downstream providers, complying with copyright laws, and publishing training data summaries27. GPAI model providers with systemic risk face additional requirements, such as performing model evaluation and reducing system ^risks27. EU Artificial Intelligence Law sets out obligations for various actors in the AI lifecycle, including suppliers, users, importers and distributors. Failure to comply with EU Artificial Intelligence Law may result in significant financial fines ²⁸. The Act will be implemented in stages: from 2 February 2025, prohibitions on the use of unacceptable risk AI systems will apply, from 2 August 2025 – obligations on general AI models, and from 2 August 2026 – obligations on high-risk AI systems. Artificial Intelligence will play a key role in overseeing the implementation and enforcement of Law ^{No 28}.

4.2. U.S. Legislation in the Field of Artificial Intelligence:

In contrast to the EU’s unified approach, AI regulation in the United States is more fragmented, with efforts at both the federal and state levels. While there is no comprehensive federal law on artificial intelligence, several legislative initiatives and executive orders address specific aspects of artificial intelligence. The National Artificial Intelligence Initiative Act of 2020 aims to promote research and developments in the field of AI ³¹. In October 2023, President Biden issued an executive order on safe, reliable, and reliable artificial intelligence, focusing on various aspects, including security standards, privacy protection, and the promotion of civil rights ³⁴. However, the current administration under President Trump is likely to adopt a softer approach to regulating the development and deployment of AI.

Several AI-related bills have been introduced in Congress, such as the AI Development and Reliability Act and the AI Creation Act, with a focus on AI safety and research ^resources.32 However, the prospects for broad, congressionally passed artificial intelligence legislation remain uncertain. In the absence of comprehensive federal regulation, states are increasingly taking the lead in passing their own laws, 32. Trends in state legislation include a focus on consumer protection and transparency, addressing the risks associated with deepfakes, and regulating the government’s use of AI tools ³³. Colorado, for example, passed the nation’s first comprehensive consumer-focused artificial intelligence law, requiring organizations to avoid algorithmic discrimination and comply with certain risk management and transparency requirements^. Utah has passed a law requiring disclosure of generative AI use, and California has a strong focus on transparency in training data. Such a decentralized approach can lead to a complex patchwork of regulations for businesses operating in different states.

4.3. Canadian Artificial Intelligence Law and Policy:

Canada is actively working to create a framework for AI regulation. The proposed Artificial Intelligence and Data Act (AIDA), introduced as part of the Digital Charter Implementation Act in 2022, aims to create a framework for the responsible design, development and deployment of AI systems that impact the lives of 36 Canadians. as Parliament has been suspended, the government remains committed to strengthening Canada’s approach to AI ³⁹. AIDA offers a risk-based framework, focusing on high-performance AI systems and establishing requirements related to human oversight, transparency, fairness, safety, accountability, credibility, and reliability.36 The Act also outlines prohibitions on the reckless and malicious use of AI that can cause serious harm, with potential criminal penalties.

In March 2025, the federal government launched Canada’s first AI strategy for the federal public sector, based on four key principles: human-centered, collaborative, prepared, and accountable. Canada is also a signatory to the Council of Europe’s Framework Convention on Artificial Intelligence and Human Rights, Democracy, and the Rule of Law, demonstrating its commitment to robust AI governance and international Cooperation. While AIDA is still under review, the government has also introduced a Voluntary Code of Conduct for the Responsible Development and Management of Advanced Generative Artificial Intelligence Systems to provide interim guidance for Canadian companies^. The proposed AIDA aims to create an AI and Data Commissioner position to monitor compliance and intervene where necessary to ensure the safety and non-discrimination of AI systems ³⁷.

4.4. Artificial Intelligence Regulation and Policy in the UK:

The United Kingdom has adopted an innovation-oriented approach to AI regulation, opting for a cross-sectoral, results-based framework based on the five pillars, rather than implementing comprehensive AI legislation at this stage. These five non-binding principles are: safety, security and reliability; appropriate transparency and explainability; fairness; accountability and governance; and contestability and redress ⁴¹. The UK Government intends to draw on the experience of existing industry regulators to interpret and apply these principles in their respective areas ⁴².

While the government has resisted legislative regulation of AI, the Artificial Intelligence (Regulation) AI Bill was reintroduced in March 2025, proposing the creation of an Artificial Intelligence Authority and the role of an AI officer for organizations involved in AI ⁴¹. However, this bill currently lacks government support and faces challenges in pushing through Law ⁴⁵The UK government has established a steering committee to support regulatory capacity and coordination in AI governance and is launching consultations on an inter-economic AI risk register and regulatory assessment^. They are also developing a voluntary AI safety standard and an AI Governance Framework framework to set minimum standards of good practice for companies selling AI products and services^. The UK’s approach emphasises flexibility and aims to balance innovation with safety by applying existing technology-neutral rules to AI42.

4.5. Regulation and policy in the field of artificial intelligence in Japan:

Japan initially took an innovative approach to AI governance, primarily relying on non-binding guidelines and strategies rather than strict regulations. In April 2024, the Ministry of Economy, Trade and Industry (METI) released the “AI for Business Guide Version 1.0”, aligned with international discussions and trends, providing voluntary guidance for developers, suppliers and users of AI ⁴⁷These guidelines emphasize a human-centred approach based on the principles of dignity, diversity and inclusion, and sustainability. ⁴⁷ They provide guidance to AI developers, AI providers and AI business users on desirable voluntary actions.

However, there is a growing recognition of the need for more formal regulation to combat emerging risks. In February 2024, the government unveiled a draft of the “Basic Law for the Promotion of Responsible AI” (AI Law) 49. This law, which is expected to be presented at the 2025 session of parliament, may include provisions for the appointment of “developers of specific models of artificial intelligence infrastructure”, who will have obligations related to security systems and reporting^. Prime Minister Fumio Kishida has said that Japan may start regulating AI in response to growing risks. It appears that Japan’s approach is evolving from a reliance on self-regulation to a more hybrid model that balances innovation with the necessary oversight, which could be affected by events in the EU and the US.

4.6. Artificial Intelligence Regulation and Policy in China:

China has adopted a more proactive and mandatory approach to AI regulation compared to many other regions 51. Since 2021, China has introduced detailed regulations on various applications of AI, including recommender algorithms and deep synthesis technologies ⁵¹. These regulations often prioritize control over information and require service providers to adhere to socialist values^. In September 2024, China released the AI Safety Management Framework, which prioritizes ethical issues in AI development, including safety, transparency, and accountability.54 The framework outlines principles such as prioritizing innovation, establishing governance mechanisms, promoting safe and reliable artificial intelligence, and protecting citizens^{‘ rights.54}

The Chinese government adopts a multi-layered and category-oriented approach to managing AI applications based on risk levels, with higher-risk use cases requiring increased control and oversight ⁵⁴. Regulations also focus on data security, protection of personal information, and the establishment of ethical review systems ⁵³. Developers, service providers, and users have specific responsibilities set out in the system, including conducting security assessments, obtaining user consent, and understanding data processing practices. China has also established an algorithm registry in which services with public opinion attributes or social mobilization capabilities are required to provide information about their algorithms . The overall approach reflects a balance between promoting AI innovation and implementing safety measures to prevent social and economic harm, which is in line with China’s ambition to become a global leader in AI by 2030.

4.7. Artificial Intelligence Regulation and Policy in Australia:

Australia currently has no legislation or regulation on AI at the federal ^level.56 However, in 2019, the government adopted eight voluntary ethical principles for AI to guide organisations in self-regulation. These principles cover human, social and environmental well-being; people-centred values; fairness; privacy protection and security; reliability and security; security; transparency and explainability; contestability and compensation for damages; and accountability ⁵⁷.

In June 2023, the Australian Government released a discussion paper on safe and responsible AI in Australia, seeking views on mitigating potential risks and supporting responsible practices ⁵⁶. The government’s interim response acknowledges that existing laws may be inadequate and indicates that it will consult on new mandatory restrictions for organizations developing and deploying AI systems in the context of high-risk. In addition, a voluntary AI security standard has been proposed, consisting of 10 voluntary safeguards applicable to AI systems of any level of risk46. These constraints focus on areas such as accountability, risk management, testing, transparency, and stakeholder engagement . In addition, since September 1, 2024, the Australian government has implemented the Responsible AI in Government Policy, which is mandatory for non-corporate Commonwealth organizations and aims to be accountable and transparent in the implementation of AI in government^. While Australia has yet to enact comprehensive legislation on artificial intelligence, there has been a clear movement towards establishing more specific regulations, especially for high-risk applications.

Table 2: Summary of key AI legislation by region

Region/Country	Legislation/Policy	General provisions	Status/Chronology
European Union	EU Artificial Intelligence Law No 28	Risk-based approach (unacceptable, high, limited, minimal risk), prohibitions on high-risk practices, obligations for high-risk AI (risk management, data governance, transparency, human control), commitments under GPAI models.	Published July 12, 2024 Phased implementation from February 2025.
United States	Disparate Federal and State Laws 31	Federal focus on research and development (NAII Act), state consumer protection laws (Colorado AI Act), deepfakes, and transparency.	Changing landscape, no comprehensive federal law. State laws vary.
Canada	Proposed Artificial Intelligence and Data Act (AIDA) 36	Risk-based framework for high-performance AI, requirements for human oversight, transparency, fairness, security, accountability. Prohibitions and punishments.	The proposed bill has not yet been adopted.
United Kingdom	Principles-based approach 41	Five cross-sectoral non-binding principles (security, transparency, fairness, accountability, adversariality). Proposed AI Regulation Bill.	First of all, soft law, the potential for future targeted legislation.
Japan	Artificial Intelligence for Business Recommendations v1.0 47	Voluntary Guidelines for AI Developers, Vendors, and Users. A potential upcoming law on artificial intelligence, with obligations for designated developers.	The guidelines are in place. The law on artificial intelligence is expected in 2025.
China	AI Safety Management System, Regulations on Algorithms and Deep Synthesis 51	Mandatory regulations on recommendation algorithms, deep synthesis, generative artificial intelligence. Multi-level risk management. Obligations to developers, providers, users.	An active regulatory environment with implemented rules.
Australia	AI Voluntary Ethics Principles 57, Proposed Mandatory Fencing 46	Eight Voluntary Ethical Principles. Proposals for mandatory fences for high-risk AI. Responsible AI policy in government agencies.	First of all, voluntary, moving towards potential mandatory rules for high-risk AI.

5. Impact of Ethical and Unethical Use of AI:

The way AI is applied, whether ethical or unethical, has profound implications for culture, society, and the workforce.

5.1. Cultural influence:

60. By respecting different cultural norms, values, and expectations, AI systems can be designed to avoid perpetuating biases that may be inherent in the data on which they are trained. However, the current situation with AI ethics often reflects Western-centric biases, which can lead to challenges in AI development. really sensitive to the nuances of non-Western cultures. Terms such as “fairness” and “privacy” can have different meanings in different cultural contexts, necessitating a more inclusive approach to AI governance.60

On the other hand, unethical AI carries the risk of distorting and stereotyping non-Western cultures, potentially leading to incorrect cultural appropriation and marginalization. For example, generative AI models have been shown to portray cultures through the lens of an outsider, sometimes reinforcing exotic or inaccurate images. and affect people’s personal identities and economic opportunities ⁶⁵. The lack of genuine cultural fluency in current AI models, which often resort to simplistic and harmful stereotypes, highlights the significant work that remains to be done in this area ⁶⁶. Without careful consideration of cultural diversity in the development of AI, there is a risk of exacerbating global cultural asymmetries and undermining trust in these technologies ⁶⁷.

5.2. Social impact:

The social impact of the ethical use of AI is overwhelmingly positive, with the promise of improving decision-making processes, increasing fairness and strengthening privacy protections ²⁶. By adhering to ethical principles, AI can be used to solve social problems, promote inclusion, and ensure that technological advances benefit all members of society. role in protecting privacy, identifying and mitigating misinformation and bias, and ultimately contributing to more equitable and just societies ⁷⁰.

On the other hand, the unethical use of AI poses significant societal risks. Biased AI systems can lead to discriminatory outcomes in critical areas such as recruitment, credit, and criminal justice, perpetuating and even exacerbating existing inequalities. Violations of privacy and human rights are another major concern, as unethical AI can be used to collect and process huge amounts of personal data without consent, potentially leading to misuse and violation of fundamental freedoms ⁷⁴. The spread of misinformation and the creation of deepfakes through unethical AI can undermine trust in sources of information and threaten democratic processes75. In addition, the lack of transparency and accountability in some AI systems can make it difficult to address harms and undermine public trust in institutions ⁶⁸. The consequences of ignoring the ethics of AI can lead to increased social injustice, loss of trust, and slower societal progress.

5.3. Clerical influence:

The application of AI has significant implications for clerical work and employment, both ethical and unethical considerations. Ethical AI can automate many of the routine and repetitive tasks typically performed in clerical positions, resulting in increased efficiency and productivity. Such automation can free up people to focus on more complex, strategic, and creative tasks, potentially leading to enriching jobs and creating new opportunities. However, the widespread adoption of AI in clerical work also raises ethical concerns about the potential displacement of jobs and the need to retrain and upskill the workforce^.

Unethical or unsupervised adoption of AI in office facilities can exacerbate socioeconomic inequalities, leading to significant job losses, especially for low-skilled workers who may find it difficult to transition to new positions. There are also concerns about the potential for increased control and oversight by employers through AI, which could negatively affect workers’ autonomy, dignity and mental health ⁸⁷. Algorithmic bias in AI systems used to perform tasks such as job allocation or performance evaluation can lead to discriminatory outcomes and unfair treatment ⁷⁶. Lack of transparency in how these AI systems make decisions can also create difficulties in determining accountability and addressing any negative impacts on workers ⁷⁶. While AI offers opportunities to streamline clerical work, it is critical to consider the ethical implications related to job safety, worker well-being, and fair treatment to ensure a responsible transition to an AI-driven workforce.

6. Future trends and forecasts:

Experts have made various predictions about the future ethical use of AI, expressing a mixture of optimism and concern. A major concern for many is that ethical principles focused primarily on the public good may not be the main driver of AI development by 2030. potentially overshadowing ethical considerations. Difficulties in reaching a global consensus on ethical standards and the challenges of enforcing them are also key challenges ⁹¹.

Despite these concerns, there is a sense of optimism about the growth of the “making AI safe” ^ecosystem.93 As new regulations are implemented and public awareness of AI ethics issues increases, organizations are expected to face more pressure to demonstrate their commitment to ethical principles. This could lead to more specialized ethical AI teams and a greater focus on third-party risk assessment.93 In addition, the development of AI ethical standards and benchmarks is expected to be critical to identifying and measuring fairness, transparency, and other key ethical characteristics of AI systems^.

Looking to the future, unethical AI will exacerbate existing social inequalities and create new ^{challenges.94} Systematic error in algorithms caused by problems in training data and design processes is a major concern, especially in sensitive areas such as healthcare and criminal justice. veracity in AI-generated ^content.95 In addition, ongoing AI-powered automation is expected to continue to have an impact on the labor market, potentially leading to job displacement and increased economic inequality.96 However, there is also hope that by focusing on the ethical development of AI, these challenges can be mitigated, leading to a future that will lead to a future that can lead to the development of AI ethically. in which AI technologies will be used responsibly for the benefit of individuals and communities by reducing bias and promoting equity 73.

Findings:

The analysis shows a global recognition of the importance of ethical considerations in the development and application of AI. Microsoft has created a robust and well-documented framework based on six core principles supported by an internal standard, specialized governance frameworks, and a wide range of resources and tools. Comparing this framework to Google, Amazon, IBM, and Meta highlights a common set of ethical principles, although the emphasis and implementation strategies may differ. The legal landscape is characterized by dynamic evolution, with the EU taking the lead in establishing comprehensive rules, while other regions are exploring different approaches. Understanding the cultural, social, and clerical implications of ethical and unethical AI is critical to addressing the challenges and reaping the benefits of this transformative technology. While ethical AI promises to bring about positive change in society, unethical applications carry significant risks related to bias, discrimination, privacy, and economic disruption. Future trends suggest a growing emphasis on AI safeguards and the emergence of ethical standards, but concerns remain about the primacy of ethical principles in the pursuit of profit and control. The comprehensive table of resources provided in this report offers a valuable framework for further in-depth exploration of these critical issues.

COMMENTS:

As can be seen from this detailed review, the governments of all countries are well aware of the need for legal regulation of AI, the establishment of uniform norms for the ethics of the use of AI and its safety.

And in Russia, legislators are actively working on these important issues. Federal Law No 258-FZ “On Experimental Legal Regimes in the Field of Digital Innovations” was adopted on July 31, 2020, several Federal Laws on amendments to the Law on Banking Activities, Government Resolutions on the establishment of EPR for unmanned vehicles in medicine.

Currently, the Law “On the Regulation of Artificial Intelligence Systems in Russia” is being developed. The group includes lawyers, representatives of consulting companies and market participants developing AI, they said. The initiative is formalized in the form of a bill “On the regulation of artificial intelligence (AI) systems in Russia.” The document was discussed at meetings of working groups in the State Duma, in Business Russia, as well as at industry discussion platforms, RBC’s interlocutors noted.

The draft law proposes:

• Introduce the concept of artificial intelligence (a set of technological and software solutions that imitate human cognitive functions, including self-learning, search for solutions without a predetermined algorithm, and provide results comparable to the results of human intellectual activity).
• The concept of AI technologies (these include technologies of computer vision, natural language processing, speech recognition and synthesis, data mining, machine learning, etc.).
• The concept of AI systems (information systems that imitate human cognitive functions using the technologies listed above), the developer, operator and user of such systems, etc.
• Introduce requirements for the labeling of AI systems, that is, a mandatory information designation that will indicate the use of AI in user interaction. Labeling must be “explicit, unambiguous and easily identifiable.” It must be provided by the operator of the AI system before each user interaction. The requirement will not apply to AI systems used in closed technological processes, for internal automation of organizations, and a number of others.
• To introduce a classification of AI systems according to the level of potential risk: systems with an unacceptable, high, limited and minimal level of risk. It is proposed to prohibit in Russia the development and operation of AI systems that belong to the category with an unacceptable level of risk, that is, that pose a threat to the security of the individual, society and the state, as well as violate the fundamental rights and freedoms of man and citizen; for high-risk systems (those used in healthcare, transport, finance, law enforcement and other areas where it can create significant risks to the life, health, fundamental rights and freedoms of citizens) to introduce mandatory state registration and certification; for those with a limited level of risk – internal quality control systems, voluntary certification.
• Introduce liability for causing harm to life, health or property for persons involved in the development and operation of artificial intelligence systems. But the document stipulates that the developer will not be liable if he has taken all necessary measures to prevent harm, if the damage occurred due to a violation of the rules for the operation of the system and in several other exceptions, the operator of AI systems will be exempt from liability if he proves that the damage occurred through the fault of the developer as a result of the user’s violation of the rules for using the system, etc high-risk AI systems.
• To determine who owns the rights to the result of intellectual activity created using AI, it is proposed to focus on whether there was a significant creative contribution of a person (whether a person determined the parameters and criteria for creating the result, whether he made a creative selection and refinement, whether he made creative decisions in the process of creation). If there was such a contribution, then the exclusive right to the result should belong to this person, if there was none, then to the operator of the AI system (valid for 50 years).

If adopted in its current form, operators and developers of AI systems will have to comply with its requirements within a year of its entry into force; high-risk AI systems that were created before the adoption of the project will have to undergo mandatory certification within two years.97

I believe that the drafters of the draft law are on the right track. By analogy with the European regulation on AI, it is proposed to classify AI systems according to the level of potential risk. Systems with an unacceptable level of risk are proposed to be banned in Russia both for operation and development. For systems with medium risk, according to the project, mandatory certification and registration are provided, and for systems with low risk – voluntary. According to experts, the new legislative initiative will launch a system for accounting for domestic solutions with AI with a risk status in order to quickly respond to potential data leaks.98

This is what is really relevant and requires the attention of legal scholars and legislators. I would like to believe in the common sense of the deputies, who must finally determine the true status of AI and AI robots as objects of law (not subjects), taking into account the modified norms of Roman law (the status of a slave).

References

1. AI Best Practices for Cloud Strategy – Cloud Adoption Framework | Microsoft Learn, accessed March 13, 2025, https://learn.microsoft.com/en-us/azure/cloud-adoption-framework/strategy/inform/ai

2. What is responsible AI – Azure Machine Learning | Microsoft Learn, accessed March 13, 2025, https://learn.microsoft.com/en-us/azure/machine-learning/concept-responsible-ai?view=azureml-api-2

3. Principles and Approach to Responsible AI | Microsoft AI, accessed March 13, 2025, https://www.microsoft.com/en-us/ai/principles-and-approach

4. Empowering Responsible AI Practices – Microsoft, accessed March 13, 2025, https://www.microsoft.com/en-us/ai/responsible-ai

5. Microsoft Principles for the Responsible Use of Artificial Intelligence | Microsoft Community Center, accessed March 13, 2025, https://techcommunity.microsoft.com/t5/ai-azure-ai-services/microsoft-responsible-ai-principles/m-p/4037307

6. What is responsible AI? – Microsoft Support, accessed March 13, 2025, https://support.microsoft.com/en-us/topic/what-is-responsible-ai-33fc14be-15ea-4c2c-903b-aa493f5b8d92

7. How to use AI responsibly and safely in Microsoft Copilot? – Service Center of the former USSR, accessed March 13, 2025, https://servicecenter.fsu.edu/s/article/How-do-I-use-AI-responsibly-and-securely-within-Microsoft-Copilot

8. Microsoft’s Commitment to Responsible AI: Leadership in Ethical AI Development, as of March 13, 2025, https://www.biggreenit.com/microsofts-responsible-ai-commitment-leading-the-way-in-ethical-ai-development/

9. Responsible AI in Action, Part 2: Take an Impact Assessment | By Kate B | Data Science at Microsoft | Medium, accessed March 13, 2025, https://medium.com/data-science-at-microsoft/responsible-ai-in-action-part-2-complete-an-impact-assessment-9b792409e8db

10. Artificial Intelligence Overview – Microsoft Service Assurance…, accessed March 13, 2025, https://learn.microsoft.com/en-us/compliance/assurance/assurance-artificial-intelligence

11. Microsoft AI Services Code of Conduct | Microsoft Learn, accessed March 13, 2025, https://learn.microsoft.com/en-us/legal/ai-code-of-conduct

12. Principles of AI – Google AI – Google AI, accessed March 13, 2025, https://ai.google/responsibility/principles/

13. AI Principles – Google AI, accessed March 13, 2025, https://ai.google/static/documents/EN-AI-Principles.pdf

14. Introduction to Responsible Artificial Intelligence | Machine Learning | Google for…, as of March 13, 2025, https://developers.google.com/machine-learning/guides/intro-responsible-ai

15. AI at Google: Our Principles, as of March 13, 2025, https://blog.google/technology/ai/ai-principles/

16. Responsible AI | Google Cloud | Google Cloud, as of March 13, 2025, https://cloud.google.com/responsible-ai

17. Responsible AI – Building AI Responsibly – AWS as of March 13, 2025, https://aws.amazon.com/ai/responsible-ai/

18. AWS Responsible AI Policy as of March 13, 2025, https://aws.amazon.com/ai/responsible-ai/policy/

19. Responsible Artificial Intelligence – Amazon Science, accessed March 13, 2025, https://www.amazon.science/tag/responsible-ai

20. AI Ethics | IBM, as of March 13, 2025, https://www.ibm.com/impact/ai-ethics

21. IBM Artificial Intelligence Pillars – IBM Policy as of March 13, 2025, https://www.ibm.com/policy/ibm-artificial-intelligence-pillars/

22. What is AI ethics? | IBM, as of March 13, 2025, https://www.ibm.com/think/topics/ai-ethics

23. ai.meta.com, as of 13 March 2025, https://ai.meta.com/static-resource/responsible-use-guide/

24. Top 10 Companies with Ethical AI Practices | AI Magazine, accessed March 13, 2025, https://aimagazine.com/ai-strategy/top-10-companies-with-ethical-ai-practices

25. Understanding AWS Responsible AI: Key Concepts and Measurements – Dojo Tutorials accessed March 13, 2025 https://tutorialsdojo.com/understanding-aws-responsible-ai-key-concepts-and-dimensions/

26. What is responsible AI? – IBM, accessed March 13, 2025, https://www.ibm.com/think/topics/responsible-ai

27. Texts of the Act | EU Law on Artificial Intelligence, as of March 13, 2025, https://artificialintelligenceact.eu/the-act/

28. The long-awaited EU Law on Artificial Intelligence becomes law after publication in the Official Journal of the EU, as of March 13, 2025, https://www.whitecase.com/insight-alert/long-awaited-eu-ai-act-becomes-law-after-publication-eus-official-journal

29. Artificial Intelligence Act, Regulation (EU) 2024/1689 – References, accessed March 13, 2025, https://www.artificial-intelligence-act.com/Artificial_Intelligence_Act_Links.html

30. Artificial Intelligence Regulatory Framework Update: Q4 2024 | Kramer Levine, accessed March 13, 2025, https://www.kramerlevin.com/en/perspectives-search/ai-regulatory-update-december-2024.html

31. Artificial Intelligence Legislation in the United States: 2025 Overview – SIG, accessed March 13, 2025, https://www.softwareimprovementgroup.com/us-ai-legislation-overview/

32. U.S. federal regulation of AI is likely to be more lenient, but states could fill the void – Skadden, as of March 13, 2025, https://www.skadden.com/insights/publications/2025/01/2025-insights-sections/revisiting-regulations-and-policies/us-federal-regulation-of-ai-is-likely-to-be-lighter

33. 3 trends emerge as AI legislation gains momentum – National…, as of March 13, 2025, https://www.ncsl.org/state-legislatures-news/details/3-trends-emerge-as-ai-legislation-gains-momentum

34. United States Approach to Artificial Intelligence – European Parliament, accessed March 13, 2025, https://www.europarl.europa.eu/RegData/etudes/ATAG/2024/757605/EPRS_ATA(2024)757605_EN.pdf

35. S.3205 – Federal Artificial Intelligence Risk Management Act 2023 118th Congress (2023-2024), accessed March 13, 2025, https://www.congress.gov/bill/118th-congress/senate-bill/3205

36. Artificial Intelligence and Data Act (AIDA) – Companion Document, accessed March 13, 2025, https://ised-isde.canada.ca/site/innovation-better-canada/en/artificial-intelligence-and-data-act-aida-companion-document

37. Artificial Intelligence Rules in Canada: An Overview of Artificial Data and …, accessed March 13, 2025, https://www.diligent.com/resources/blog/canada-ai-regulations

38. Artificial Intelligence and Data Act – Innovation, Science and Economic Development Canada, accessed March 13, 2025, https://ised-isde.canada.ca/site/innovation-better-canada/en/artificial-intelligence-and-data-act

39. Recent developments in AI in federal government agencies …, accessed March 13, 2025, https://www.nortonrosefulbright.com/en/knowledge/publications/01b26022/recent-developments-on-ai-in-federal-government-institutions

40. Global AI Governance Law and Policy: Canada – IAPP, accessed March 13, 2025, https://iapp.org/resources/article/global-ai-governance-canada/

41. Global AI Governance Law and Policy: UK – IAPP, accessed 13 March 2025, https://iapp.org/resources/article/global-ai-governance-uk/

42. AI Regulatory System in the UK – Deloitte, accessed March 13, 2025, https://www.deloitte.com/uk/en/Industries/financial-services/blogs/the-uks-framework-for-ai-regulation.html

43. Overview of AI regulations in the UK – Diligent, accessed March 13, 2025, https://www.diligent.com/resources/blog/uk-ai-regulation

44. AI Watch: A Global Regulatory Tracker – Russia | White & Case LLP, as of March 13, 2025, https://www.whitecase.com/insight-our-thinking/ai-watch-global-regulatory-tracker-united-kingdom

45. Artificial Intelligence (Regulation) Bill: Closing the AI Regulatory Gap in the UK?, accessed 13 March 2025, https://kennedyslaw.com/en/thought-leadership/article/2025/the-artificial-intelligence-regulation-bill-closing-the-uks-ai-regulation-gap/

46. Shaping the Future: Australia’s Approach to AI Regulation…, accessed 13 March 2025, https://www.technologyslegaledge.com/2024/09/shaping-the-future-australias-approach-to-ai-regulation/

47. Artificial Intelligence Regulations in Japan – Diligent, accessed March 13, 2025, https://www.diligent.com/resources/blog/japan-ai-regulations

48. Technology – AI Governance in Japan: An Overview of Current Policies and Future Possible Regulations, as of March 13, 2025, https://grjapan.com/sites/default/files/content/articles/files/20241115%20GR%20Japan%20Industry%20Insight%20AI%20in%20Japan_5.pdf

49. Understanding AI Regulations in Japan – Current State and Future…, as of March 13, 2025, https://www.dlapiper.com/en/insights/publications/2024/10/understanding-ai-regulations-in-japan-current-status-and-future-prospects

50. Pragmatic Model of Artificial Intelligence Management in Japan – The Diplomat, accessed March 13, 2025, https://thediplomat.com/2025/02/japans-pragmatic-model-for-ai-governance/

51. Tracing the Origins of AI Regulation in China | Carnegie Endowment for International Peace, accessed March 13, 2025, https://carnegieendowment.org/2024/02/27/tracing-roots-of-china-s-ai-regulations-pub-91815

52. Artificial Intelligence Regulations in China and How They Are Created, as of March 13, 2025, https://carnegieendowment.org/2023/07/10/china-s-ai-regulations-and-how-they-get-made-pub-90117

53. Making Sense of AI Rules in China – Holistic AI, accessed March 13, 2025, https://www.holisticai.com/blog/china-ai-regulation

54. China-Releases-AI-Security-governance-framework | DLA Piper, as of March 13, 2025, https://www.dlapiper.com/en/insights/publications/2024/09/china-releases-ai-safety-governance-framework

55. Managing AI in China: Strategies, Initiatives, and Key Considerations – Bird & Bird, accessed March 13, 2025, https://www.twobirds.com/en/insights/2024/china/ai-governance-in-china-strategies-initiatives-and-key-considerations

56. Australia | Herbert Smith Freehills | Global Law Firm, accessed March 13, 2025, https://www.herbertsmithfreehills.com/insights/reports/searchlight-ai/australia

57. Principles of AI Ethics in Australia – Department of Industry, Science and Resources, accessed March 13, 2025, https://www.industry.gov.au/publications/australias-artificial-intelligence-ethics-principles/australias-ai-ethics-principles

58. Responsible AI Policy in Government | aga – Australian Government Architecture, accessed March 13, 2025, https://architecture.digital.gov.au/responsible-use-of-AI-in-government

59. AI in Public Policy | digital.gov.au, as of March 13, 2025, https://www.digital.gov.au/policy/ai/policy

60. Global Ethics of AI: Bridging Cultural Differences in Technology…, accessed March 13, 2025, https://technologymagazine.com/articles/global-ai-ethics-bridging-cultural-divides-in-technology

61. ISACA Now Blog 2024 AI Ethics Naviging Different Cultural Contexts, accessed March 13, 2025, https://www.isaca.org/resources/news-and-trends/isaca-now-blog/2024/ai-ethics-navigating-different-cultural-contexts

62. The Cultural Impact of AI: Shaping Society and Identity | by Megasis Network | Medium, accessed March 13, 2025, https://megasisnetwork.medium.com/the-cultural-impact-of-ai-shaping-society-and-identity-03af47bebd41

63. How Culture Shapes What People Want From AI | Stanford HAI, accessed March 13, 2025, https://hai.stanford.edu/news/how-culture-shapes-what-people-want-ai

64. AI and the cultural impact of AI technologies – Franki T, accessed March 13, 2025, https://www.francescatabor.com/articles/2024/8/3/ai-and-the-cultural-impact-of-ai-technologies

65. Non-Western cultures are distorted and harmed by generative AIs…, as of March 13, 2025, https://ist.psu.edu/about/news/venkit-aies-social-harms

66. Does Artificial Intelligence Have a Culture Problem?, accessed March 13, 2025, https://news.theculturefactor.com/news/does-artificial-intelligence-have-a-culture-problem

67. Global perspectives on AI bias: addressing cultural asymmetries and ethical implications, accessed 13 March 2025, https://www.orfonline.org/expert-speak/global-perspectives-on-ai-bias-addressing-cultural-asymmetries-and-ethical-implications

68. Ethical and Social Implications of AI | Princeton Review, accessed March 13, 2025, https://www.princetonreview.com/ai-education/ethical-and-social-implications-of-ai-use

69. AI Inventions – Ethical and Social Implications | Intellectual Property Management, accessed March 13, 2025, https://www.managingip.com/article/2bc988k82fc0ho408vwu8/expert-analysis/ai-inventions-the-ethical-and-societal-implications

70. Artificial Intelligence, Ethics and Social Responsibility – PECB, accessed March 13, 2025, https://pecb.com/article/artificial-intelligence-ethics-and-social-responsibility

71. AI Ethics: What It Is, Why It Matters, and More – Coursera, accessed March 13, 2025, https://www.coursera.org/articles/ai-ethics

72. ETHICAL AND SOCIAL IMPLICATIONS OF AI – IEEE CAI 2023, accessed March 13, 2025, https://cai.ieee.org/2023/ethics/

73. Ethical and Unethical AI: Bridging the Gap – Codoid Innovations, accessed March 13, 2025, https://codoid.com/ai/ethical-and-unethical-ai-bridging-the-divide/

74. Ethical AI: The Implications of Unethical AI | MorphCast Blog, accessed March 13, 2025, https://www.morphcast.com/blog/ethical-ai-the-potential-consequences-of-unethical-ai-in-2023/

75. Uncovering the Ethical Challenges of AI: Building a Fair, Transparent, and Responsible Future, accessed March 13, 2025, https://arize.com/blog/ai-ethical-issues-unraveled-building-a-fair-transparent-and-responsible-future/

76. (PDF) ETHICAL IMPLICATIONS OF AI IN BUSINESS – ResearchGate accessed March 13, 2025, https://www.researchgate.net/publication/385782217_ETHICAL_IMPLICATIONS_OF_AI_IN_BUSINESS

77. Ethical Aspects of AI: Fairness, Transparency and Framework | The Future of Responsible AI | Lumenalta, as of March 13, 2025, https://lumenalta.com/insights/ethical-considerations-of-ai

78. community.trustcloud.ai, accessed 13 March 2025, https://community.trustcloud.ai/docs/grc-launchpad/grc-101/risk-management/risks-and-consequences-of-irresponsible-ai-in-organizations-the-hidden-dangers/#:~:text=The%20consequences%20of%20irresponsible%20AI,-The%20implementation%20of&text=One%20of%20the%20major%20consequences,to%20unfair%20treatment%20or%20discrimination.

79. 14 Risks and Dangers of Artificial Intelligence (AI) – Built-in, as of March 13, 2025, https://builtin.com/artificial-intelligence/risks-of-artificial-intelligence

80. Risks and consequences of irresponsible AI in organizations: hidden dangers, accessed March 13, 2025, https://community.trustcloud.ai/docs/grc-launchpad/grc-101/risk-management/risks-and-consequences-of-irresponsible-ai-in-organizations-the-hidden-dangers/

81. AI Ethics, Part One: Overcoming the Pressure for Responsible AI – Alvarez and Marsal, accessed March 13, 2025, https://www.alvarezandmarsal.com/insights/ai-ethics-part-one-navigating-pressures-responsible-ai

82. Vision | Dangers of irresponsible AI – Dataiku Knowledge Base, accessed March 13, 2025, https://knowledge.dataiku.com/latest/ml-analytics/responsible-ai/concept-dangers-irresponsible-ai.html

83. Topics – Social risks of generative AI – Charity Digital, accessed March 13, 2025, https://charitydigital.org.uk/topics/the-social-risks-of-generative-ai-11000

84. Ethical Implications of AI and Job Displacement – Sogeti Labs, accessed March 13, 2025, https://labs.sogeti.com/the-ethical-implications-of-ai-and-job-displacement/

85. What are the positive and negative impacts of AI on employment? – IoT Academy, accessed March 13, 2025, https://www.theiotacademy.co/blog/ai-impact-on-employment/

86. AI is unlikely to take over most jobs, but clerical work faces the risk of automation: UN study, accessed March 13, 2025, https://www.gadgets360.com/internet/news/ai-artificial-intelligence-unlikely-destroy-most-jobs-clerical-workers-at-risk-ilo-4318148

87. Artificial Intelligence: Ethics, Governance and Regulation – UK Parliament POST, accessed March 13, 2025, https://post.parliament.uk/artificial-intelligence-ethics-governance-and-regulation/

88. The Ethics of Generative AI: 10 Ethical Challenges (with Best Practices) – eWEEK, accessed March 13, 2025, https://www.eweek.com/artificial-intelligence/generative-ai-ethics/

89. Generative AI, the American Worker, and the Future of Work – Brookings Institution, accessed March 13, 2025, https://www.brookings.edu/articles/generative-ai-the-american-worker-and-the-future-of-work/

90. The Impact of Artificial Intelligence on Human Society and Bioethics – PMC, accessed March 13, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC7605294/

91. Survey XII: What is the future of ethical AI design? | Presenting…, as of March 13, 2025, https://www.elon.edu/u/imagining/surveys/xii-2021/ethical-ai-design-2030/

92. Ethical Applications of Artificial Intelligence: Evidence from Veterans’ Health Studies, accessed March 13, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC8209529/

93. Our predictions for ethical AI in 2022 – Credo AI News, accessed March 13, 2025, https://www.credo.ai/news/our-predictions-for-ethical-ai-in-2022

94. Health Equity and Ethical Considerations in the Use of Artificial Intelligence in Public Health and Medicine – CDC, accessed March 13, 2025, https://www.cdc.gov/pcd/issues/2024/24_0245.htm

95. Never assume that the accuracy of AI information equals the truth, as of March 13, 2025, https://unu.edu/article/never-assume-accuracy-artificial-intelligence-information-equals-truth

96. Ethical challenges are growing as AI takes on a larger role in decision-making – Harvard Gazette, accessed March 13, 2025, https://news.harvard.edu/gazette/story/2020/10/ethical-concerns-mount-as-ai-takes-bigger-decision-making-role/

97. Russia will discuss a ban on AI with a “threatening level of risk” — RBC https://www.rbc.ru/technology_and_media/11/04/2025/67f7dc399a79477fdd97bf30

98. AI Regulation: From Sandboxes to Standards, https://pravo.ru/story/258257/

1. Gemini is a powerful artificial intelligence (AI) model from Google that can understand text, images, video, and audio. As a multimodal model, Gemini is described as capable of performing complex tasks in mathematics, physics, and other fields, as well as understanding and generating high-quality code in a variety of programming languages. ↑
2. Scientific Platform ” XXI CENTURY” journal, # 1, April 2025, pp. 4-13, www.episs.ru ↑
3. Translation by Microsoft Translation service. ↑
4. Meta Corporation is recognized as an extremist organization in Russia. ↑
5. Meta Corporation is recognized as an extremist organization in Russia. ↑
6. Meta Corporation is recognized as an extremist organization in Russia. ↑