Presentations in this session:

• Presentation by Lisa Talia Moretti: Why we need responsible AI practitioners more than ever?

With the launch of the EU's AI Act as well as other guidelines and principles launched by other member states, the AI legislative landscape has undoubtedly improved. However, in the race to implement AI one of the things that seems to have happened is that compliance has superseded ethics and responsible AI practice. Responsible AI practitioners have become humans at risk in the AI race, ironically, during a time when we're needed most. Blinded by marketing language and unfounded promises, many organisations are buying products and services that are unsafe, open them up to cyber and privacy risks, have failure rates that should never be accepted and sometimes, aren't AI at all. The risks are multidimensional and should be unacceptable. This talk will cover (1) the reasons ethics and compliance are both needed for responsible AI practices to thrive, (2) the difference between compliance and ethics, (3) spotlight benefits of each and (4) outline a day in the life of a responsible AI practitioner. The audience who attends this talk will be exposed to the benefits of employing responsible AI practitioners, what skills they should look out for when hiring them and the various roles and responsibilities they can fulfil within an organisation.

• Presentation by Omer Shafiq, CEO, Hovi ID

As AI agents begin to interact on behalf of organizations, establishing trust becomes essential. Individuals need to know not just who they are engaging with, but also whether that agent is authorized to act on behalf of an organization, and under what conditions. This presentation explores how organizational digital wallets, combined with verifiable credentials (VCs) and decentralized identifiers (DIDs), can enable AI agents to prove their identity, authority, and intent in real time. Built on self-sovereign identity (SSI) principles and aligned with eIDAS 2.0 and the European Digital Identity Wallet (EUDI) framework, this model creates a foundation for trustworthy, auditable, and privacy-respecting digital interactions. For the MyData audience, this session offers a forward-looking perspective on how automation can be made verifiable and accountable by design. Attendees will see how organizations can hold and issue credentials through their own wallets, enabling machine agents to act transparently and within defined limits. Just as individuals must be empowered with control over their data, institutions too must operate within ethical and verifiable frameworks. Rather than treat AI as a black box, we can give it a cryptographically verifiable identity rooted in human-centric values. This presentation shows how trust in automation, organizational responsibility, and data sovereignty can coexist through open, interoperable identity infrastructure.

• Presentation by Sneha Das, Assistant Professor, Technical University of Denmark: AI Safety, Alignment, and (the Myth of Isolated) Privacy: Reframing Data Agency in the Age of General-Purpose AI (GPAI)

This talk challenges the prevailing notion that data privacy is an individual concern, arguing instead that privacy, AI safety, and alignment are deeply interdependent. As GPAI become more embedded in everyday life, their alignment with diverse human intents grows both more critical and more complex. We reflect on the 'no free lunch' principle to show how alignment is shaped and constrained by the training data itself, often reflecting systemic biases and consent asymmetries. Individual opt-outs offer limited protection when data infrastructures remain misaligned with collective values. We explore: (a) why AI safety must go beyond preventing catastrophic failures to ensuring everyday alignment with human agency; (b) how training data built on unequal participation undermines fairness and pluralism; and (c) why privacy must be reframed as a collective alignment issue, not just a personal choice. Combining technical insights, real-world examples, and policy implications, this talk offers the MyData community a framework for understanding data agency in the age of AI. Attendees will leave with a clearer view of how consent, control, and co-governance must evolve to address the collective stakes of data-driven systems.
 

This session is hosted by Sneha Das, Assistant Professor, Technical University of Denmark