AIER Summary
Topic 2: Regulation of AI & the AI Act...................................................................................... 3
General Context of AI Regulation.......................................................................................3
Paving the Way for the AI Act............................................................................................ 4
The AI Act itself.................................................................................................................. 4
Sanctions and Enforcement............................................................................................... 8
Regulatory Sandboxes....................................................................................................... 8
Topic 3: Ethics and the law.......................................................................................................9
Relationship Between Law and Ethics............................................................................... 9
Interaction of Ethics and Law............................................................................................. 9
Weaknesses or Difficulties in Relation to Ethics................................................................ 9
Branches of Ethics........................................................................................................... 10
Main Normative Theories................................................................................................. 10
Principlism........................................................................................................................ 11
Is AI Ethically Neutral?..................................................................................................... 12
Topic 4: Liability......................................................................................................................13
Setting the scene..............................................................................................................13
Types of Liability...............................................................................................................13
General Challenges in AI Liability.................................................................................... 13
User Liability: Autonomous Vehicles as a Case Study.....................................................14
Producer Liability: The Product Liability Directive (PLD)..................................................15
Topic 5: AI & Data Protection (GDPR)................................................................................... 17
Introduction to GDPR in the Context of AI........................................................................17
Scope of Application of the GDPR to AI Systems............................................................ 17
Key GDPR Principles and their Application to AI............................................................. 18
Data Subject Right in the Context of AI............................................................................19
Challenges and Real-World Examples.............................................................................20
Topic 6: Explainable AI (XAI)..................................................................................................21
What is Explainable AI (XAI)?.......................................................................................... 21
Why is XAI Needed?........................................................................................................ 21
Technical Aspects of XAI..................................................................................................21
Legal Aspects of XAI........................................................................................................ 22
Challenges and Considerations....................................................................................... 23
Topic 7: AI & Human Enhancement....................................................................................... 25
Defining Human Enhancement (HE)................................................................................ 25
Examples of Human Enhancement Technologies (HETs)................................................25
Ethics of Human Enhancement........................................................................................26
Current Regulatory Framework........................................................................................ 27
Key Points and Open Legal Questions............................................................................ 28
Topic 8: AI & Law Enforcement.............................................................................................. 29
Contextualizing Law Enforcement....................................................................................29
Traditional Policing vs. Predictive Policing....................................................................... 29
Controversial Smart Technologies....................................................................................30
, Legal Challenges..............................................................................................................30
Under-Regulated Areas....................................................................................................32
Topic 9: AI & Media................................................................................................................ 33
Challenges of AI in Media................................................................................................ 33
EU Regulatory Solutions.................................................................................................. 34
Media Organizations and AI Act Compliance...................................................................35
Topic 10: AI Regulation and Technical Standards..................................................................36
What are Technical Standards?....................................................................................... 36
Why are Standards Important?........................................................................................ 36
Standards under the EU's New Legislative Framework (NLF)......................................... 36
Standards in the AI Act.....................................................................................................37
Challenges and Criticisms................................................................................................38
Topic 11: AI and Intellectual Property (IP).............................................................................. 40
Intellectual Property 101...................................................................................................40
AI and Copyright Law....................................................................................................... 40
What AI Technology Can (and Cannot) Be Copyrighted?................................................ 40
Copyright Protection for AI-Generated Output................................................................. 40
AI Training Data and Copyright........................................................................................ 41
AI and Patent Law............................................................................................................ 42
Patentability of AI Technology.......................................................................................... 42
Patentability of AI-Generated Output (Inventorship)........................................................ 43
Conclusion........................................................................................................................44
,Topic 2: Regulation of AI & the AI Act
General Context of AI Regulation
The necessity and approach to AI regulation are complex and multifaceted, raising several key
questions:
● Why Regulate AI? The rapid advancement of AI, particularly since 2018 with
developments like Google's BERT, OpenAI's DALL-E and ChatGPT, and Google's
Gemini, has led to concerns about potential risks. Stephen Hawking, for instance, voiced
worries about superintelligent AI surpassing human control and leading to unintended
consequences, including existential risks. Regulation aims to ensure that AI development
remains beneficial to humanity, preventing it from "turning evil" and ensuring that
corporate profit is not the sole guiding principle. There's a constant tension between
fostering innovation and protecting fundamental rights and safety.
● What to Regulate? Legislators must decide whether to regulate the technology itself or
focus on the consequences of its use. The AI Act largely adopts a consequence-based,
risk-oriented approach.
● When to Regulate? Regulation can be ex-ante (preventive, before harm occurs, e.g.,
transparency obligations) or ex-post (remedial, after harm, e.g., liability rules). The AI Act
incorporates both.
● At Which Level? AI is a global phenomenon, necessitating consideration of regulation at
international, regional (EU), and national levels.
○ International Initiatives: Various international organizations have attempted to
introduce norms and guidelines since 2018, though binding international treaties
are difficult due to the need for agreement among diverse countries. The Council
of Europe, for example, adopted a framework convention on AI in 2023 to ensure
consistency with human rights, democracy, and the rule of law.
○ EU's Broader Regulatory Landscape: Beyond the AI Act, the EU has other
instruments regulating new technologies, including the General Data Protection
Regulation (GDPR), the Digital Services Act (DSA), the Digital Markets Act
(DMA), and the Data Act.. These existing laws are still applicable and must be
complied with alongside the AI Act.
○ Global Approaches:
■ China: Has been active in regulating AI since 2021, with provisions on
algorithmic recommendations (promoting "positive values"), personal
information protection (similar to GDPR), deep synthesis (labeling
deepfakes), and generative AI (aligning with "core socialist values" and
censorship).
, ■ South Korea: Approved the AI Basic Act in December 2024, influenced
by the EU AI Act with a risk-based classification but generally less
prescriptive.
■ US: Lacks a comprehensive federal AI act, instead relying on executive
orders (e.g., maintaining leadership, safe/secure/trustworthy
development) and state-level regulations (e.g., AI in recruitment). These
executive orders can be revoked by subsequent administrations, leading
to policy shifts.
Paving the Way for the AI Act
The AI Act did not emerge in a vacuum but was the culmination of a gradual process of EU
engagement with AI:
● Early EU Approach (2018): Focused on staying ahead of technological developments,
encouraging AI uptake, preparing for socio-economic changes, and establishing an
appropriate ethical and legal framework.
● High-Level Expert Group on AI (HLEG): Established in 2018, this group produced
non-binding Ethics Guidelines for Trustworthy AI and an Assessment List for Trustworthy
AI (ALTAI). These guidelines, while not legally binding, served as a foundational basis for
the AI Act, translating ethical norms into practical questions for developers to assess
risks and mitigate harms. They cover seven key requirements: human agency and
oversight, transparency, diversity/non-discrimination/fairness, societal/environmental
well-being, and accountability.
● EU Commission White Paper on AI (2020): This paper laid the foundation for the EU's
regulatory approach, proposing a risk-based framework for trustworthy AI and discussing
the need for a new legal framework or adjustments to existing laws.
● Relationship between Law and Ethics: This preparatory phase highlighted the intrinsic
link. Both law and ethics are normative, guiding behavior and promoting decent conduct.
Ethical values often form the foundation for law, with the HLEG guidelines being a prime
example for the AI Act. However, law is binding and enforceable by the state, while ethics
is not. Ethics alone lacks enforcement mechanisms and can be too generic or abstract
for practical implementation by technical professionals.
The AI Act itself
The AI Act was formally proposed by the European Commission in 2021. After negotiations
between the European Parliament and the Council, it was adopted on March 13, 2024, and
entered into force on August 1, 2024, though most rules will apply later, typically by August 2,
2026.
● Definition of an AI System: The AI Act defines an AI system as a "machine-based
system that is designed to operate with varying levels of autonomy and that, for explicit or
implicit objectives, infers, from the input it receives, how to generate outputs such as
predictions, content, recommendations, or decisions that can influence physical or virtual
Topic 2: Regulation of AI & the AI Act...................................................................................... 3
General Context of AI Regulation.......................................................................................3
Paving the Way for the AI Act............................................................................................ 4
The AI Act itself.................................................................................................................. 4
Sanctions and Enforcement............................................................................................... 8
Regulatory Sandboxes....................................................................................................... 8
Topic 3: Ethics and the law.......................................................................................................9
Relationship Between Law and Ethics............................................................................... 9
Interaction of Ethics and Law............................................................................................. 9
Weaknesses or Difficulties in Relation to Ethics................................................................ 9
Branches of Ethics........................................................................................................... 10
Main Normative Theories................................................................................................. 10
Principlism........................................................................................................................ 11
Is AI Ethically Neutral?..................................................................................................... 12
Topic 4: Liability......................................................................................................................13
Setting the scene..............................................................................................................13
Types of Liability...............................................................................................................13
General Challenges in AI Liability.................................................................................... 13
User Liability: Autonomous Vehicles as a Case Study.....................................................14
Producer Liability: The Product Liability Directive (PLD)..................................................15
Topic 5: AI & Data Protection (GDPR)................................................................................... 17
Introduction to GDPR in the Context of AI........................................................................17
Scope of Application of the GDPR to AI Systems............................................................ 17
Key GDPR Principles and their Application to AI............................................................. 18
Data Subject Right in the Context of AI............................................................................19
Challenges and Real-World Examples.............................................................................20
Topic 6: Explainable AI (XAI)..................................................................................................21
What is Explainable AI (XAI)?.......................................................................................... 21
Why is XAI Needed?........................................................................................................ 21
Technical Aspects of XAI..................................................................................................21
Legal Aspects of XAI........................................................................................................ 22
Challenges and Considerations....................................................................................... 23
Topic 7: AI & Human Enhancement....................................................................................... 25
Defining Human Enhancement (HE)................................................................................ 25
Examples of Human Enhancement Technologies (HETs)................................................25
Ethics of Human Enhancement........................................................................................26
Current Regulatory Framework........................................................................................ 27
Key Points and Open Legal Questions............................................................................ 28
Topic 8: AI & Law Enforcement.............................................................................................. 29
Contextualizing Law Enforcement....................................................................................29
Traditional Policing vs. Predictive Policing....................................................................... 29
Controversial Smart Technologies....................................................................................30
, Legal Challenges..............................................................................................................30
Under-Regulated Areas....................................................................................................32
Topic 9: AI & Media................................................................................................................ 33
Challenges of AI in Media................................................................................................ 33
EU Regulatory Solutions.................................................................................................. 34
Media Organizations and AI Act Compliance...................................................................35
Topic 10: AI Regulation and Technical Standards..................................................................36
What are Technical Standards?....................................................................................... 36
Why are Standards Important?........................................................................................ 36
Standards under the EU's New Legislative Framework (NLF)......................................... 36
Standards in the AI Act.....................................................................................................37
Challenges and Criticisms................................................................................................38
Topic 11: AI and Intellectual Property (IP).............................................................................. 40
Intellectual Property 101...................................................................................................40
AI and Copyright Law....................................................................................................... 40
What AI Technology Can (and Cannot) Be Copyrighted?................................................ 40
Copyright Protection for AI-Generated Output................................................................. 40
AI Training Data and Copyright........................................................................................ 41
AI and Patent Law............................................................................................................ 42
Patentability of AI Technology.......................................................................................... 42
Patentability of AI-Generated Output (Inventorship)........................................................ 43
Conclusion........................................................................................................................44
,Topic 2: Regulation of AI & the AI Act
General Context of AI Regulation
The necessity and approach to AI regulation are complex and multifaceted, raising several key
questions:
● Why Regulate AI? The rapid advancement of AI, particularly since 2018 with
developments like Google's BERT, OpenAI's DALL-E and ChatGPT, and Google's
Gemini, has led to concerns about potential risks. Stephen Hawking, for instance, voiced
worries about superintelligent AI surpassing human control and leading to unintended
consequences, including existential risks. Regulation aims to ensure that AI development
remains beneficial to humanity, preventing it from "turning evil" and ensuring that
corporate profit is not the sole guiding principle. There's a constant tension between
fostering innovation and protecting fundamental rights and safety.
● What to Regulate? Legislators must decide whether to regulate the technology itself or
focus on the consequences of its use. The AI Act largely adopts a consequence-based,
risk-oriented approach.
● When to Regulate? Regulation can be ex-ante (preventive, before harm occurs, e.g.,
transparency obligations) or ex-post (remedial, after harm, e.g., liability rules). The AI Act
incorporates both.
● At Which Level? AI is a global phenomenon, necessitating consideration of regulation at
international, regional (EU), and national levels.
○ International Initiatives: Various international organizations have attempted to
introduce norms and guidelines since 2018, though binding international treaties
are difficult due to the need for agreement among diverse countries. The Council
of Europe, for example, adopted a framework convention on AI in 2023 to ensure
consistency with human rights, democracy, and the rule of law.
○ EU's Broader Regulatory Landscape: Beyond the AI Act, the EU has other
instruments regulating new technologies, including the General Data Protection
Regulation (GDPR), the Digital Services Act (DSA), the Digital Markets Act
(DMA), and the Data Act.. These existing laws are still applicable and must be
complied with alongside the AI Act.
○ Global Approaches:
■ China: Has been active in regulating AI since 2021, with provisions on
algorithmic recommendations (promoting "positive values"), personal
information protection (similar to GDPR), deep synthesis (labeling
deepfakes), and generative AI (aligning with "core socialist values" and
censorship).
, ■ South Korea: Approved the AI Basic Act in December 2024, influenced
by the EU AI Act with a risk-based classification but generally less
prescriptive.
■ US: Lacks a comprehensive federal AI act, instead relying on executive
orders (e.g., maintaining leadership, safe/secure/trustworthy
development) and state-level regulations (e.g., AI in recruitment). These
executive orders can be revoked by subsequent administrations, leading
to policy shifts.
Paving the Way for the AI Act
The AI Act did not emerge in a vacuum but was the culmination of a gradual process of EU
engagement with AI:
● Early EU Approach (2018): Focused on staying ahead of technological developments,
encouraging AI uptake, preparing for socio-economic changes, and establishing an
appropriate ethical and legal framework.
● High-Level Expert Group on AI (HLEG): Established in 2018, this group produced
non-binding Ethics Guidelines for Trustworthy AI and an Assessment List for Trustworthy
AI (ALTAI). These guidelines, while not legally binding, served as a foundational basis for
the AI Act, translating ethical norms into practical questions for developers to assess
risks and mitigate harms. They cover seven key requirements: human agency and
oversight, transparency, diversity/non-discrimination/fairness, societal/environmental
well-being, and accountability.
● EU Commission White Paper on AI (2020): This paper laid the foundation for the EU's
regulatory approach, proposing a risk-based framework for trustworthy AI and discussing
the need for a new legal framework or adjustments to existing laws.
● Relationship between Law and Ethics: This preparatory phase highlighted the intrinsic
link. Both law and ethics are normative, guiding behavior and promoting decent conduct.
Ethical values often form the foundation for law, with the HLEG guidelines being a prime
example for the AI Act. However, law is binding and enforceable by the state, while ethics
is not. Ethics alone lacks enforcement mechanisms and can be too generic or abstract
for practical implementation by technical professionals.
The AI Act itself
The AI Act was formally proposed by the European Commission in 2021. After negotiations
between the European Parliament and the Council, it was adopted on March 13, 2024, and
entered into force on August 1, 2024, though most rules will apply later, typically by August 2,
2026.
● Definition of an AI System: The AI Act defines an AI system as a "machine-based
system that is designed to operate with varying levels of autonomy and that, for explicit or
implicit objectives, infers, from the input it receives, how to generate outputs such as
predictions, content, recommendations, or decisions that can influence physical or virtual
