AI ETHICS AND REGULATION
Contents
Lecture 1: Introduction to AIER.........................................................................................................2
Some essential (basic) information....................................................................................................2
“Regulating” AI..................................................................................................................................3
Lecture 2: Regulation of AI & the AI Act............................................................................................4
Regulating AI: General.......................................................................................................................4
The AI Act...........................................................................................................................................5
Lecture 3: Ethics and the Law...........................................................................................................8
Law vs Ethics......................................................................................................................................8
Ethical theories – 3 mainstream theories..........................................................................................9
Lecture 4: Liability and AI – Challenges in Tort Law........................................................................10
Setting the scene on liability............................................................................................................10
General challenges in Tort Law and AI.............................................................................................10
Human Machine Interaction (Case study autopilot)........................................................................12
Alternative liability regimes and challenges.....................................................................................12
Some recommendations..................................................................................................................13
Lecture 5: AI and the GDPR.............................................................................................................13
Setting the scene.............................................................................................................................13
Scope of Application of GDPR..........................................................................................................14
General principles............................................................................................................................15
Data subjects’ rights........................................................................................................................16
Lecture 6: Explainable AI – Interplay technology & law..................................................................17
Technology and explainable AI........................................................................................................17
Legal sources and XAI......................................................................................................................18
Legal Sources Beyond GDPR and AI Act...........................................................................................18
Lecture 7: AI and Human Enhancement..........................................................................................19
Part I: Definitional issues.................................................................................................................19
Part II: Societal and Ethical Issues....................................................................................................20
Part III: Current regulatory framework............................................................................................22
Lecture 8: AI and Law Enforcement................................................................................................23
Contextualizing Law Enforcement...................................................................................................23
Legal Challenges..............................................................................................................................24
Lecture 9: AI and Media – Legal Perspective...................................................................................27
The Right to Freedom of Expression (FoE).......................................................................................27
1
, Freedom of Expression & Content Moderation...............................................................................28
Freedom of Expression & AI.............................................................................................................29
Lecture 10: AI Regulation and Technical Standards.........................................................................30
Recap AI Act.....................................................................................................................................30
Role of Standards under AI Act........................................................................................................31
Lecture 11: AI and IP.......................................................................................................................32
AI & Copyright..................................................................................................................................33
AI & Patent Law...............................................................................................................................34
Attachment: Relevant Articles........................................................................................................35
Lecture 1: Introduction to AIER
Some essential (basic) information
Law= the system of rules which a particular country or community recognizes as regulating the
actions of its members and which it may enforce by the imposition of penalties.
“Make sure the state functions.”
- Public vs. Private (e.g. contract/property law)
Legal Subject= who/what is subject to the law
Distinction between:
- Law in books (written rules)
- Law in action (how rules are applied in practice)
Law: Hierarchy (basic)
- International Treaties
Importance
- European Union law
- Constitution
- Ordinary laws
Hierarchy of norms: Constitution > Laws > Decrees.
Regulations: Directly applicable
Directives: Transposed into national law, countries can choose how they do it.
Council of Europe: Broader than the EU, includes human rights frameworks.
Council of Europe (CoE) European Union (EU)
Includes all EU countries + others Only European countries meeting specific
Membership
(e.g., UK, Turkey) economic & political criteria
2
, Council of Europe (CoE) European Union (EU)
Human rights, democracy, rule of Economic & political integration, single
Main Focus
law market
Key Legal European Convention on Human Binding regulations, directives, and decisions
Instruments Rights (ECHR), soft law (CFREU)
European Court of Human Rights
Judicial Body European Court of Justice (CJEU)
(ECtHR)
Role in AI Ethical and human rights-based Binding legislative framework (e.g., AI Act,
Regulation guidance (e.g., CAHAI) GDPR)
AI and human rights reports, soft
AI-related law principles. Ensure AI activities AI Act, Digital Services Act, GDPR, Data
Initiatives are compliant with democracy and Governance Act
the rule of law.
Influence Type Normative and ethical influence Legal and regulatory enforcement
Interaction with Promotes ethical frameworks Creates binding obligations for EU states and
AIER applicable beyond the EU developers
EU legal principles:
Subsidiarity: Local level unless EU action is more effective.
“Only act if it has added value”: when a member state cannot act itself (e.g. environmental
protection)
Conferral: The EU can only act within the powers given to it by member states.
(e.g not on educational)
Proportionality: Measures must be proportionate to aims.
“Must not exceed what is necessary to achieve”:
“Regulating” AI
Regulation is not limited to laws. Instead, it involves steering behavior through multiple mechanisms
(see table below).
Modalities of regulation
Each modality: pros and cons.
In AI context: Interaction!
e.g., Gun violence regulation
- law: could do background checks
- norms: anti-gun campaigns
3
, - architecture/technology: built-in tracker
- market: increase taxes
Technology and society: different theories
Perspective Key Ideas Examples
Technological Tech shapes society; neutral &
Printing press, smartphone
Determinism inevitable
AI automation: impacting employment
Neutral + Opportunities, impact human
structures
Autonomous behavior (regulation).
Instrumentalism Tech is a neutral tool to achieve a
Nuclear energy, Social Media, AI
Neutral + Human- goal; humans define use (“dual use”:
systems, guns
controlled good or bad)
Tech embodies values, shapes
Substantivism
outcomes.
Value-driven + Guns, Internet, AI
Use not relevant, their existence is
Autonomous
reshaping society
Critical Theory Facial recognition
(=Combination of - May underrepresent minorities
Tech reflects human values; society racial/gender bias: not
previous ones)
shapes design. Autonomy user. neutral
Value-driven + - Prioritize security/control >
human-controlled privacy: value-driven
Lecture 2: Regulation of AI & the AI Act
Regulating AI: General
Regulation= steering behaviour, not just passing laws
2 approaches:
- Ex-ante: regulate before damage occurs
- Ex-post: regulate after harm happens
What are we regulating?
- The AI system itself?
- Its deployment and effects/consequences?
- The actors involved? (developers, deployers, importers, etc...)
At which level?
- The more countries involved harder negotiations less binding
International Perspectives
4
, Country/Region Initiatives
China Rules on recommendation algorithms, facial recognition, and generative AI
South Korea AI Basic Act – strong focus on governance and ethics. Classifies AI according to
(SKAIA) risk, human oversight, privacy...
No single AI law, but several sector-specific and state-level frameworks (e.g.,
United States
NIST risk management)
The AI Act
- Before it was proposed- Foundations:
o Ethics Guidelines for Trustworthy AI by the High-Level Expert Group (HLEG)
o Assessment List for Trustworthy AI (ALTAI)
o EU White Paper on AI (2020):
Ecosystem of trust (regulation)
Ecosystem of excellence (innovation)
- Goals of the AI Act:
o Encourage trustworthy AI
o Protect fundamental rights
o Promote AI innovation
o Ensure human-centric AI
“What is AI?”
- Machine-based system
- Designed to operate with varying levels of autonomy
- May exhibit adaptiveness after deployment
- Infers, from received input, how to generate outputs (predictions, content,
recommendations, or decisions)
- That can influence physical or virtual environments
Material State-of-the-Art: to the current state of technological best practices that must be
followed when designing and deploying AI, especially high-risk systems.
5 Important Exceptions (AI Act – Art. 5)
Art.5: These are AI practices that are prohibited under the Act — unless a specific exception
applies, often in strictly limited conditions:
Domain Possible Exception Context
AI use in research settings may be permitted under strict ethical and data
Research & Innovation
governance conditions, even for sensitive applications.
Exceptions may exist when AI is used in the context of providing or
Consumer access /
improving essential goods and services (e.g., credit scoring, benefit
Economic services
eligibility), though high-risk obligations apply.
Home & Private AI applications in the private/home context (e.g., smart assistants, personal
Sphere health devices) may be exempt if used non-professionally.
Security / Law Exceptions for real-time biometric identification in public spaces exist in
5
Contents
Lecture 1: Introduction to AIER.........................................................................................................2
Some essential (basic) information....................................................................................................2
“Regulating” AI..................................................................................................................................3
Lecture 2: Regulation of AI & the AI Act............................................................................................4
Regulating AI: General.......................................................................................................................4
The AI Act...........................................................................................................................................5
Lecture 3: Ethics and the Law...........................................................................................................8
Law vs Ethics......................................................................................................................................8
Ethical theories – 3 mainstream theories..........................................................................................9
Lecture 4: Liability and AI – Challenges in Tort Law........................................................................10
Setting the scene on liability............................................................................................................10
General challenges in Tort Law and AI.............................................................................................10
Human Machine Interaction (Case study autopilot)........................................................................12
Alternative liability regimes and challenges.....................................................................................12
Some recommendations..................................................................................................................13
Lecture 5: AI and the GDPR.............................................................................................................13
Setting the scene.............................................................................................................................13
Scope of Application of GDPR..........................................................................................................14
General principles............................................................................................................................15
Data subjects’ rights........................................................................................................................16
Lecture 6: Explainable AI – Interplay technology & law..................................................................17
Technology and explainable AI........................................................................................................17
Legal sources and XAI......................................................................................................................18
Legal Sources Beyond GDPR and AI Act...........................................................................................18
Lecture 7: AI and Human Enhancement..........................................................................................19
Part I: Definitional issues.................................................................................................................19
Part II: Societal and Ethical Issues....................................................................................................20
Part III: Current regulatory framework............................................................................................22
Lecture 8: AI and Law Enforcement................................................................................................23
Contextualizing Law Enforcement...................................................................................................23
Legal Challenges..............................................................................................................................24
Lecture 9: AI and Media – Legal Perspective...................................................................................27
The Right to Freedom of Expression (FoE).......................................................................................27
1
, Freedom of Expression & Content Moderation...............................................................................28
Freedom of Expression & AI.............................................................................................................29
Lecture 10: AI Regulation and Technical Standards.........................................................................30
Recap AI Act.....................................................................................................................................30
Role of Standards under AI Act........................................................................................................31
Lecture 11: AI and IP.......................................................................................................................32
AI & Copyright..................................................................................................................................33
AI & Patent Law...............................................................................................................................34
Attachment: Relevant Articles........................................................................................................35
Lecture 1: Introduction to AIER
Some essential (basic) information
Law= the system of rules which a particular country or community recognizes as regulating the
actions of its members and which it may enforce by the imposition of penalties.
“Make sure the state functions.”
- Public vs. Private (e.g. contract/property law)
Legal Subject= who/what is subject to the law
Distinction between:
- Law in books (written rules)
- Law in action (how rules are applied in practice)
Law: Hierarchy (basic)
- International Treaties
Importance
- European Union law
- Constitution
- Ordinary laws
Hierarchy of norms: Constitution > Laws > Decrees.
Regulations: Directly applicable
Directives: Transposed into national law, countries can choose how they do it.
Council of Europe: Broader than the EU, includes human rights frameworks.
Council of Europe (CoE) European Union (EU)
Includes all EU countries + others Only European countries meeting specific
Membership
(e.g., UK, Turkey) economic & political criteria
2
, Council of Europe (CoE) European Union (EU)
Human rights, democracy, rule of Economic & political integration, single
Main Focus
law market
Key Legal European Convention on Human Binding regulations, directives, and decisions
Instruments Rights (ECHR), soft law (CFREU)
European Court of Human Rights
Judicial Body European Court of Justice (CJEU)
(ECtHR)
Role in AI Ethical and human rights-based Binding legislative framework (e.g., AI Act,
Regulation guidance (e.g., CAHAI) GDPR)
AI and human rights reports, soft
AI-related law principles. Ensure AI activities AI Act, Digital Services Act, GDPR, Data
Initiatives are compliant with democracy and Governance Act
the rule of law.
Influence Type Normative and ethical influence Legal and regulatory enforcement
Interaction with Promotes ethical frameworks Creates binding obligations for EU states and
AIER applicable beyond the EU developers
EU legal principles:
Subsidiarity: Local level unless EU action is more effective.
“Only act if it has added value”: when a member state cannot act itself (e.g. environmental
protection)
Conferral: The EU can only act within the powers given to it by member states.
(e.g not on educational)
Proportionality: Measures must be proportionate to aims.
“Must not exceed what is necessary to achieve”:
“Regulating” AI
Regulation is not limited to laws. Instead, it involves steering behavior through multiple mechanisms
(see table below).
Modalities of regulation
Each modality: pros and cons.
In AI context: Interaction!
e.g., Gun violence regulation
- law: could do background checks
- norms: anti-gun campaigns
3
, - architecture/technology: built-in tracker
- market: increase taxes
Technology and society: different theories
Perspective Key Ideas Examples
Technological Tech shapes society; neutral &
Printing press, smartphone
Determinism inevitable
AI automation: impacting employment
Neutral + Opportunities, impact human
structures
Autonomous behavior (regulation).
Instrumentalism Tech is a neutral tool to achieve a
Nuclear energy, Social Media, AI
Neutral + Human- goal; humans define use (“dual use”:
systems, guns
controlled good or bad)
Tech embodies values, shapes
Substantivism
outcomes.
Value-driven + Guns, Internet, AI
Use not relevant, their existence is
Autonomous
reshaping society
Critical Theory Facial recognition
(=Combination of - May underrepresent minorities
Tech reflects human values; society racial/gender bias: not
previous ones)
shapes design. Autonomy user. neutral
Value-driven + - Prioritize security/control >
human-controlled privacy: value-driven
Lecture 2: Regulation of AI & the AI Act
Regulating AI: General
Regulation= steering behaviour, not just passing laws
2 approaches:
- Ex-ante: regulate before damage occurs
- Ex-post: regulate after harm happens
What are we regulating?
- The AI system itself?
- Its deployment and effects/consequences?
- The actors involved? (developers, deployers, importers, etc...)
At which level?
- The more countries involved harder negotiations less binding
International Perspectives
4
, Country/Region Initiatives
China Rules on recommendation algorithms, facial recognition, and generative AI
South Korea AI Basic Act – strong focus on governance and ethics. Classifies AI according to
(SKAIA) risk, human oversight, privacy...
No single AI law, but several sector-specific and state-level frameworks (e.g.,
United States
NIST risk management)
The AI Act
- Before it was proposed- Foundations:
o Ethics Guidelines for Trustworthy AI by the High-Level Expert Group (HLEG)
o Assessment List for Trustworthy AI (ALTAI)
o EU White Paper on AI (2020):
Ecosystem of trust (regulation)
Ecosystem of excellence (innovation)
- Goals of the AI Act:
o Encourage trustworthy AI
o Protect fundamental rights
o Promote AI innovation
o Ensure human-centric AI
“What is AI?”
- Machine-based system
- Designed to operate with varying levels of autonomy
- May exhibit adaptiveness after deployment
- Infers, from received input, how to generate outputs (predictions, content,
recommendations, or decisions)
- That can influence physical or virtual environments
Material State-of-the-Art: to the current state of technological best practices that must be
followed when designing and deploying AI, especially high-risk systems.
5 Important Exceptions (AI Act – Art. 5)
Art.5: These are AI practices that are prohibited under the Act — unless a specific exception
applies, often in strictly limited conditions:
Domain Possible Exception Context
AI use in research settings may be permitted under strict ethical and data
Research & Innovation
governance conditions, even for sensitive applications.
Exceptions may exist when AI is used in the context of providing or
Consumer access /
improving essential goods and services (e.g., credit scoring, benefit
Economic services
eligibility), though high-risk obligations apply.
Home & Private AI applications in the private/home context (e.g., smart assistants, personal
Sphere health devices) may be exempt if used non-professionally.
Security / Law Exceptions for real-time biometric identification in public spaces exist in
5
