The European AI Office is calling for expert evaluators to participate in an online workshop addressing systemic risks posed by general-purpose AI models. Scheduled for December 13, 2024, this exclusive event will bring together leaders in AI evaluation to refine methodologies and foster collaboration in line with the EU AI Act.
Advance AI Safety Through Collaborative Evaluation
This workshop is a critical initiative by the European AI Office to enhance the safety, trustworthiness, and accountability of general-purpose AI. Participants will share insights, discuss challenges, and showcase expertise to help build robust frameworks for systemic risk assessment.
Topics of Focus:
- CBRN Risks: AI-related threats involving chemical, biological, radiological, and nuclear risks.
- Cyber Offense: Mitigating AI-driven cyberattacks and security breaches.
- Major Accidents: Preventing disruptions to critical infrastructure or large-scale systems.
- Loss of Control: Addressing oversight issues in autonomous AI systems.
- Discrimination: Ensuring AI-generated outcomes are fair and unbiased.
- Privacy Infringements: Protecting data integrity and preventing misuse.
- Disinformation: Combating harmful or misleading AI-generated content.
- Other Systemic Risks: Addressing risks to public health, democracy, and safety.
The workshop offers a unique platform for evaluators, researchers, and organizations to contribute directly to the development of best practices for AI risk management.
Submit Your Expertise
The European AI Office invites eligible participants to submit abstracts of previously published work on systemic risk evaluations. Contributions should address one or more of the workshop’s focus areas and demonstrate a commitment to advancing AI safety.
Eligibility Criteria:
- Registered organizations or university-affiliated research groups.
- Demonstrated expertise in evaluating general-purpose AI models.
Submissions will be evaluated based on:
- Technical Quality: The depth and rigor of the research.
- Relevance: Alignment with EU AI Act priorities.
- Mission Alignment: Support for the European AI Office’s objectives.
Key Dates to Remember
- Submission Deadline: December 8, 2024 (Anywhere on Earth).
- Invitation Notification: December 11, 2024.
- Workshop Date: December 13, 2024, at 14:00 CET.
Why Join This Initiative?
Under the EU AI Act, providers of general-purpose AI models are required to evaluate and mitigate risks, ensure cybersecurity, conduct adversarial testing, and report incidents. The European AI Office enforces these requirements, ensuring compliance while fostering innovation.
Participating in this workshop allows you to:
- Shape AI Governance: Contribute to Europe’s leadership in responsible AI.
- Share Best Practices: Showcase your expertise in systemic risk evaluation.
- Collaborate with Experts: Network with industry leaders and the AI Office.
- Strengthen AI Oversight: Help develop a reliable framework for safe and ethical AI systems.
