PS 4th week

• set of principles that help guide the design, development, deployment and use of AI

  -> building trust in AI solutions

• involves the consideration of a broader societal impact of AI

• required to align thsese technologies with Stakehodler values and legal standards and ethical principles

• resp AI embeds this etical principals in AI applications and workflows to mitigate risks and negative outcomes while maximizing positive outcomes

• inclusive growth, ssustainable develoopment and well being

• transparency and explainability (factors determining unemployment rate)

• accountability

• Human rights and democratic values, including fairness and privacy

• robustness, security and safety

“Fairness is not automatic — it’s intentional”

• Diverse training datasets

• Bias audits (e.g., aequitas)

• Human-in-the-loop review

• Transparent documentation of

  model logic (e.g., Model Cards)

AI act (regulation)

• The first EU legal framework to adress the risks on AI. It sets clear requirements and obligations for AI developers and users, and positions Europe as a global leader. It also aims to reduce administrative and financial burdens, especially for SMEs.

GDPR (Regulation)

• The General Data Protection Regulation (GDPR) is an EU law governing the collection, processing, and storage of personal data. It protects individuals’ rights to privacy while allowing free data flow within the EU. Organizations must follow strict rules on transparency, consent, accountability, and security, with heavy penalties for non-compliance.

• The AI Act aims to provide AI developers and deployers with clear requirements and

  obligations regarding specific uses of AI.

prohibited AI practives:

• manipulative or deceptive techniques

• exploits any vulnerability of a natural person

• untargeted scraping of facial images for facial recognition databases

• infer emotions at workplace

Example:

AI systems intended to be used for the recruitment or selection of natural persons, (…) to analyse and filter job applications, and to evaluate candidates;

1. Bioentrics

2. Critical infrastructure

3. Education and vocational training

4. Employment, workers’ management and

   access to self-employment

5. Access to and enjoyment of essential private

   services and essential public services and benefits

6. Law enforcement

7. Migration, asylum and border control management

8. Administration of justice and democratic processes

1. risk management System

2. Data and Data governance

3. technical dodumentation

4. record keeping

5. transparency and provision of information to deployers

6. human oversight

7. accuracy, robustness and cybersecurity

8. fundamental rights impact assesement for high risk AI systems

GDPR + AI Act + other data/ AI/ digital legislation + industry specific legilation

• Legal Agreement between Service Provider & Partner/ Client

• Individual NDA

• Data Processing Agreements/ Joint Controller Agreements (GDPR)

What?

• It is a regulation in EU law on data protection and privacy in the European Union (EU) and the European Economic Area (EEA). It also addresses the transfer of personal data outside the EU and EEA areas.

• To give control to individuals over their personal data and to simplify the regulatory environment for international business by unifying the regulation within the EU.

• California Consumer Privacy Act (CCPA) has many similarities

• processed in a lawful and transparent manner

• specific purpose for processing or new purpose compatible with original one

• only necessary data -> data minimisation

• keep data up-to-date

• only user for initial purpose -> purpose limitation principle

• storage no langer than necessary -> storage limitation

• technical and organisational safeguards

• who

• why

• legal justification

• how long

• transfered outside the EU?

• copy of the data

• right to lodge a complaint

• right to withdraw consent

• existence of automated decision-making

• if core acticity is processing data on a large scale

• large scale, regular and systematic monitoring of individuals

• public administration

1. How can I measure the impact of the data and AI tool created on my original problem/ opportunity?

2. What key indicators can be connected to my tool?

3. How does performance compare before and after implementing the model?

4. Are there secondary indicators (cost, time, quality, customer satisfaction) that also reflect impact?

1. How are end-users (internal teams or customers)

experiencing the AI tool?

2. Does it enhance decision-making quality or

improve the customer experience?

3. Are there any complaints, trust issues, or ethical

concerns arising from its use?

4. Are employees actually using the AI tool in their

workflows?

5. What are the barriers to adoption across teams?

6. Is the model reducing manual effort or decision

time? If so, by how much?