The AI Act
in Five Steps
Ahcène Boubekki
Camilla MV Arffmann
Line KH Clemmensen
Sneha Das
Boris Düder
Rasmus SB Jensen
Olga Kokoulina
UCPH
UCPH
UCPH/DTU
DTU
UCPH
DTU
UCPH Law
AI Act Workshop
The Workshop
DATE
ATTENDANCE
OBJECTIVE
DATA
AI Act
New Legislative Framework
1. Law
2. Standards
Already regulated domains (Annex I)
8 high-risk domains (Annex III)
High-risk vs the Rest
In-effect Requirements
A5: Prohibited practices (manipulation...)
A4: AI Literacy
GPAI obligations (testing, report...)
A50: transparency (inform, watermark...)
Applicability
Role
Classification
Obligations
Implementation
Applicability
Applicability
Articles
Art. 3: Definitions of AI, GPAI, etc.
Art. 2 (3): Scope and Exceptions
|
|
(1) ‘AI system’ means a machine-based system that is designed to operate with varying levels of autonomy and that may exhibit adaptiveness after deployment, 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 environments. |
|
|
(63) ‘general-purpose AI model’ means an AI model, including where such an AI model is trained with a large amount of data using self-supervision at scale, that displays significant generality and is capable of competently performing a wide range of distinct tasks regardless of the way the model is placed on the market and that can be integrated into a variety of downstream systems or applications, except AI models that are used for research, development or prototyping activities before they are placed on the market; |
Generally assumed
Good Thing!
Applicability
Role
Classification
Obligations
Implementation
Role
Role
Articles
Art. 3: Definitions
Absent from the materials
Obvious? Unfamiliar?
(3) ‘provider’ means a natural or legal person, public authority, agency or other body that develops an AI system or a general-purpose AI model or that has an AI system or a general-purpose AI model developed and places it on the market or puts the AI system into service under its own name or trademark, whether for payment or free of charge;
(4) ‘deployer’ means a natural or legal person, public authority, agency or other body using an AI system under its authority except where the AI system is used in the course of a personal non-professional activity;
(6) ‘importer’ means a natural or legal person located or established in the Union that places on the market an AI system that bears the name or trademark of a natural or legal person established in a third country;
(7) ‘distributor’ means a natural or legal person in the supply chain, other than the provider or the importer, that makes an AI system available on the Union market;
Art. 16: Obligations of providers of high-risk AI systems
Art. 23: Obligations of importers
Art. 24: Obligations of distributor
Art. 26: Obligations of deployers of high-risk AI systems
Art. 25: Responsibilities along the AI value chain
Most were providers (developers)
Applicability
Role
Classification
Obligations
Implementation
Classification
Classification
Articles
Art. 3: Definitions
(2) ‘risk’ means the combination of the probability of an occurrence of harm and the severity of that harm;
Art. 6(1): Safety component or whole of a product under Annex I
Annex I
List of Union harmonisation legislation
machinery, toys, craft and watercraft, lifts, explosive protective equipments, radio equipments, pressure equipments, cablesways, personal protective equipments, fuels, medical devices, in vitro diagnostics
Annex III
High-risk AI systems referred to in Article 6(2)
Biometrics, Critical infrastructure, employment, essential services, law enforcement, migration, justice
Art. 6(2): AI systems referred to in Annex III
Art. 6(3): Narrow procedural task
Under 6 (1) high-risk is not automatic!
risks to health, safety, and fundamental rights?
GPAI
Query
Main Agent
agent
Memory
Planner
Censor
Output
Web
Tools
Scripts
etc.
agent
agent
Main Agent
Preparatory Work
Improve results
Narrow procedural task
Control
the AI system is intended to detect decision-making patterns or deviations from prior decision-making patterns and is not meant to replace or influence the previously completed human assessment, without proper human review; or
Classification
Articles
Art. 3: Definitions
Art. 6(1): Safety component or whole of a product under Annex I
Art. 6(2): AI systems referred to in Annex III
Art. 6(3): Narrow procedural task
(4) 'profiling' means any form of automated processing of personal data consisting of the use of personal data to evaluate certain personal aspects relating to a natural person, in particular to analyse or predict aspects concerning that natural person's performance at work, economic situation, health, personal preferences, interests, reliability, behaviour, location or movements;
(4) 'profiling' means any form of automated processing of personal data consisting of the use of personal data to evaluate certain personal aspects relating to a natural person, in particular to analyse or predict aspects concerning that natural person's performance at work, economic situation, health, personal preferences, interests, reliability, behaviour, location or movements;
No exception if PROFILING!
Human Involvement does not matter!
All based on
intended purpose
(12) ‘intended purpose’ means the use for which an AI system is intended by the provider, including the specific context and conditions of use, as specified in the information supplied by the provider in the instructions for use, promotional or sales materials and statements, as well as in the technical documentation;
More intuition than rules...
Tried to escape with
human in the loop
Applicability
Role
Classification
Obligations
Implementation
Obligations
Obligations
Articles
Art. 4: Literacy
Art. 50: Transparency
Art. 9-15,17,43,72,73: for high risk
High-Risk Paths:
Annex I: Follow sector procedure and report to sector authtorities
Annex III: self assessment, register in EU database,
if filter, justify!
in some cases: risk to fund. rights assessment.
High-risk = Conformity assessment
FOR ALL AI SYSTEMS!
Uncertainty among participants
Overlooked AI literacy!
...bias due to format...
Frictions
Legal vs Engineers
Applicability
Role
Classification
Obligations
Implementation
Implementation
Implementation
Articles
Art. 40: Harmonized Standards
Which metric?
Which tools?
Threshold or
just reporting?
Standards = Presumption of Conformity