Did you miss #AIAthens2024? Was there a session or speaker that you want to revisit? Recordings of the event are now available to view on our YouTube channel:

Our 3-part series covers:.⚖️ 1: What are red lines & why they matter.🏛️ 2: Real-world examples already in practice.🤝 3: Pathways to effective global red lines. 📘 Read the full series: #AIRedLines #AIGovernance #AIPolicy

⚠️ Without firm red lines, AI systems risk causing widespread, cross-border harm. Current dynamic rewards speed over safety. We need governance that scales with technology.

Red lines aren't theoretical:. 🛡️ EU bans unacceptable uses like social scoring and manipulation. 🛡️ U.S. states prohibit unlawful deep fakes.

🚧 Worried about AI going unchecked? The current dynamic rewards speed of AI development over safety. But we can choose a different way forward. Today we published "Global Red Lines for AI," a 3-part explainer series on boundaries that AI systems must not cross. Thread 👇

The prospect of AI agents is exciting. But without proper rules, they could challenge rights & safety. The EU should address open questions regarding regulation of agents, & other jurisdictions should follow its approach. 📘 Read the report: [5/5].

For each mechanism, we map out specific obligations per actor:. 🧠 Model Providers.🛠 Agent System Developers.🚦 System Deployers. No single actor has all the information, access, or expertise—governance must be shared across the AI value chain. [4/5]

How does the AI Act govern agents at the level of General Purpose-AI models & in specific applications?. We identify 10 ways to govern agents across 4 pillars, citing AI Act articles:. 1️⃣ Risk assessments. 2️⃣ Transparency tools . 3️⃣ Technical controls . 4️⃣ Human oversight . [3/5]

We just released the first deep dive into how the EU AI Act applies to AI agents. Our report breaks it down:. ✅ Where the Act already applies. ⚠️ Where it falls short. 🔧 What governance mechanisms are needed. [2/5]

AI agents are here. Going beyond chatbots, agents can plan, browse, code, and interact with the real world—all without human approval. But the big question is are we ready to govern them?. [1/5].

RT @Forum_Global: 📸 Session 5a at #AI2025USA where the panel discussed ensuring #AI innovation, cooperation & security. Speakers were:.‣@b….

RT @RyanFedasiuk: President Trump's gathering of world leaders in Qatar isn't business-as-usual. From gutting compute controls to striking….

Learn more and apply by May 9:

Are you an expert in AI policy looking to help shape the future of #AIGovernance in Europe? 🇪🇺 . We’re #hiring an experienced policy leader with an established network and credibility within the European AI governance ecosystem to work with us in Brussels.

Efficient operations drive impact 🚀 Seeking a seasoned COO to lead human resources, finance, administration, systems, & compliance so our nonprofit organization can level up to better advance AI governance. Applications are open until April 18. Rolling review, so apply soon! ⤵️.

How can the EU Code of Practice for General-Purpose AI Model Providers realize the promise of the #AIAct, by promoting innovation & protecting citizens? Read our preliminary analysis & five recommendations for improvement of the third draft of the Code:

@NicolasMoes @elespanolcom @e_paniagua What commitments would you like to see from your political leaders and AI companies after the #AIActionSummit? Let us know ⬇️.

@NicolasMoes @elespanolcom As Moës says, “We welcome the voluntary commitments made so far, but we must go beyond goodwill gestures.” . Read @e_paniagua's full article:

The voluntary commitments released after the #AIActionSummit by industry leaders can only go so far. As our Executive Director @NicolasMoes notes in @elespanolcom, the lack of enforcement mechanisms “jeopardize the likelihood of tangible impact.”.

This kind of collaboration is what we need to build a movement across civil society, which can then better work together with government, industry, and academia to build the robust AI governance ecosystem we urgently need. 🤝🌍.