Corporate leadership across member states faces direct regulatory scrutiny as courts assign full statutory liability for automated outputs to the deploying enterprise. Customising…
14 septembre 2026 – mis à jour il y a 1 jour
Publications étiquettées avec "agentic ai"
“Uranium-Grade Control over AI Agents Is Achievable.”
By “uranium-grade,” I use the term purely as an analogy for exceptionally high control rigor — a very high level of technical control…
14 septembre 2026 – mis à jour il y a 1 jour
AI systems are increasingly capable of initiating consequential actions in digital and physical environments: transferring funds, altering records, sending binding…
14 septembre 2026 – mis à jour il y a 1 jour
As AI systems evolve from conversational tools into increasingly agentic systems capable of retrieving information, coordinating specialised components, proposing decisions and initiating…
13 septembre 2026 – mis à jour il y a 1 jour
Yesterday, I came across the recent OpenAI “wiki incident”, and it gave me a very different perspective on the governance challenges emerging with multi-agent AI.
It…
9 septembre 2026 – mis à jour il y a 5 jours
An Architecture for Technical Enforcement for GDPR and the EU AI Act in Critical AI SYSTEMS
Your agent can move money, ship code, export a customer database, or…
7 septembre 2026 – mis à jour il y a 5 jours
The AI Act establishes how artificial intelligence should be developed and deployed. Agentic AI introduces an additional requirement: organizations must be able to define, enforce and prove the…
- 2 septembre 2026 – mis à jour il y a 1 semaine
As AI systems become increasingly agentic, a new question is emerging.
We are already asking what an agent is authorised to do, which tools it may use,…
28 août 2026 – mis à jour il y a 2 semaines
Anthropic's benchmark deployment running an autonomous AI swarm across a complex research problem resulted in $100,000 in compute token consumption within 36 hours. While computer…
- 20 août 2026 – mis à jour il y a 3 semaines
Dear members of the Apply AI Alliance,
As the European AI Office advances the drafting of the Codes of Practice and prepares the operational guidelines for the EU AI Act, the rapid…
30 juillet 2026 – mis à jour il y a 1 mois