Capabilities and Challenges of AI in Digital Governance
summary
Reuters-backed EU benchmark finds major models miss AI Act requirements: ChatGPT 49% compliance, Alibaba 39%, prompt-hijacking resistance shows Meta ~42% vs Google ~85%. Examples include a GPT miscounting Greek letters in a name. Describes Nvidia GPUs as the current backbone, a government BI project using data analytics for school HR allocation, Ministry of Education AI assistance, and procurement fraud detection, and an Athens competence center working with Microsoft, IBM, SAP, Snowflake, Databricks and Demokritos to build internal responsible-AI tooling.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 1:10AI Act compliance challenges
- 5:14AI evolution to generative models
- 8:50AI disrupting industries
- 12:26AI in digital governance
- 14:59Generative AI limitations and risks
- 16:47Responsible AI framework
- 17:53AI platform initiatives
- 19:45AI in Greek government projects
- 20:30Q&A
related talks
01h 09mGOVERNANCE AND REGULATION OF AI FROM THE PERSPECTIVE OF AUTONOMY AND PRIVACY
28mEthical AI?!
34m“You Can’t Just Turn It On”: The Hidden Work Behind Enterprise AI
13mAI as a Strategic Asset: Building Competitive Advantage
26mLeadership in Times of AI
45mPanel: Rethinking Skills and Work in the Age of AI
01h 08mTHE GOVERNANCE OF AI: CONVERGENCE OR DIVERGENCE?
01h 14mGenerative AI in Academia Navigating Challenges and Opportunities...