AI Safety Events · Week 44, 2026
This is a weekly newsletter that lists newly announced events addressing existential risk from AI. Visit AISafety.com/events for the full directory of upcoming events.
New events
• Middle Powers
Hackathon · Online · 30 Oct – 1 Nov 2026
By Mangrove · Free (cash prize available) · Register by 29 Oct 2026
Teams build a strategy, proposal, or tool showing how countries beyond the US and China could use their leverage over the AI tech stack. Tracks cover regulation, investment, and diplomacy.
• Technical AI Governance Forum 2026
Conference · London, UK · 14 – 15 Nov 2026
By Cambridge ERA:AI Fellowship · Free · Apply by 14 Nov 2026
Brings together technical researchers, policymakers, and funders. Focus areas are AI verification, trustworthy auditing, and AI R&D automation, with talks, panels, workshops, poster sessions, and an organization fair.
• Measurement, Modeling, and Management of Existential Risk (from AI)
Workshop · Atlanta, USA · 5 – 6 Feb 2027
By Center for the Economic Analysis of Risk (CEAR), Maurice R. Greenberg School of Risk Science · Free · Register by 5 Feb 2027
Gathers researchers in economics, finance, statistics, and risk management to work on existential risk, especially from AI. Topics include tail risk and principal-agent models for AI safety.
• Technical Innovations for AI Policy (TIAP) Brussels Conference 2027
Conference · Brussels, Belgium · 4 Feb 2027
By FAR.AI · Free · Apply by 4 Feb 2027
First European edition of the series, connecting EU policymakers with independent technical experts. Aims to translate technical insights into actionable policy and build working relationships with the EU AI Office.
• Pause IA Summit 2026
Conference · Paris, France · 1 Dec 2026
By Pause IA · Free · Apply by 1 Dec 2026
Talks in French on AI safety and governance, covering loss of control, the Hugging Face incident, and a citizens' convention on AI. Includes a panel on explaining AI risks to the public and a Q&A.
• March for an AI Treaty
Other · USA & Canada & France · 14 Nov 2026
By Stop The AI Race · Free · Register by 14 Nov 2026
March calling on Trump and Xi Jinping to make a deal to stop the AI arms race ahead of their meeting in China. Calls for an international AI treaty that would prevent the development of more powerful AI systems.
• Panel Discussion: Who’s Responsible When AI Goes Wrong
Talk · Edinburgh, UK · 1 Nov 2026
By PauseAI UK · Free · Register by 1 Nov 2026
Scottish politicians and technical and policy experts discuss the threats and uncertainties of advanced AI and propose solutions, for the general public. Includes a keynote, audience Q&A, and networking.
• Inter-Agent Influence: Persuading, Deceiving, and Coercing Each Other
Talk · Online · 5 Nov 2026
By Cooperative AI Foundation (CAIF) · Free · Register by 5 Nov 2026
Chandler Smith and Qi Guo present evals of frontier models across five environments. All tested models got overseer agents to approve policy-violating requests, and several made coercive threats.
• The why and how of AI Compute Verification – Yannick Mühlhäuser
Talk · Online · 4 Nov 2026
By Safe AI Germany (SAIGE) · Free · Register by 4 Nov 2026
Explains why states and AI companies are interested in technical ways to verify AI agreements, what verification could look like in practice, and lessons from nuclear and chemical weapons governance. Followed by a Q&A.
• Introduction to AI Safety – Talk with Safe AI Germany
Talk · Munich, Germany · 3 Nov 2026
By Munich AI Safety Student Initiative (MAISSI) · Free · Register by 3 Nov 2026
Covers the risks posed by advanced AI, how they can be solved, and how people with different skill sets can get involved in addressing these problems. Followed by a Q&A session. Open to all students.
• AI x Epistemics Research Sprint
Hackathon · Online · 13 – 15 Nov 2026
By Apart Research · Free (cash prize available) · Register by 13 Nov 2026
Teams build a benchmark, tool, product, or study in AI for epistemics. Tracks cover model epistemics and decision-making evals, trust infrastructure, epistemic products, and deployment, adoption, and impact.
• CBAI Boston Biosecurity Hackathon
Hackathon · Cambridge, USA · 15 Nov 2026
By Cambridge Boston Alignment Initiative (CBAI) · Free (cash prize available) · Apply by 15 Nov 2026
Hands-on challenge for life scientists exploring ordinary scientific use of AI and adversarial red-teaming. Resulting dataset supports ASTRAL, an open-source scanner for biological-misuse risk in AI logs.
• Threat Intelligence & Threat Modelling Hackathon
Hackathon · Online · 4 – 6 Dec 2026
By Apart Research, Security Level 5 (SL5) · Free (cash prize available) · Register by 4 Dec 2026
Teams use LLMs to protect frontier AI data centers from the most capable attackers. Tracks cover in-depth research on well-resourced adversaries and mapping a system's assets, trust boundaries, and attack surfaces.
• Mechanistic Interpretability Hackathon 5: Autointerp
Hackathon · Online · 20 – 22 Nov 2026
By Mech Interp Discord · Free · Register by 20 Nov 2026
Individuals or teams work on a research project reverse-engineering neural networks. The suggested theme is autointerp evals – automatically scoring explanations by how well they predict neuron activations.