Upcoming events

Annual mechanistic interpretability workshop at ICML 2026 in Seoul. Focuses on developing principled methods to analyze and understand model internals - weights and activations. Brings together researchers from academia, industry, and independent research. CFP deadline May 8 (AoE). Follows successful previous editions at ICML 2024 and NeurIPS 2025.

{
  "type_weight": 0.85,
  "source_trust": 0.9,
  "topic_relevance": 0.95,
  "time_proximity": 0.7433962264150944,
  "community_signal": 0.85,
  "speaker_org_signal": 0.85,
  "is_cfp_open": 1,
  "source_count": 2
}

Workshop at ICML 2026 exploring pluralistic AI: aligning with the diversity of human values. Accepts 4-8 page papers plus unlimited references. Topics span machine learning, philosophy, HCI, social sciences, and policy on methods for pluralistic ML training, value conflict handling, and approaches to diverse societal values. CFP deadline May 3, camera-ready June 10. Non-archival format accepting position papers, works in progress, policy papers, and academic papers.

{
  "type_weight": 0.85,
  "source_trust": 0.85,
  "topic_relevance": 0.8,
  "time_proximity": 0.7383647798742139,
  "community_signal": 0.7,
  "speaker_org_signal": 0.75,
  "is_cfp_open": 1,
  "source_count": 2
}

Anthropic's four-month fellowship program for AI safety research. Weekly stipend of $3,850 USD / £2,310 GBP / $4,300 CAD, plus ~$15k/month compute budget and close mentorship from Anthropic researchers. Priority areas include scalable oversight, adversarial robustness, and interpretability. Application deadline May 3.

{
  "type_weight": 0.8,
  "source_trust": 0.95,
  "topic_relevance": 0.95,
  "time_proximity": 0.6930817610062893,
  "community_signal": 0.85,
  "speaker_org_signal": 0.95,
  "is_cfp_open": 1,
  "source_count": 1
}

Large-scale summit hosted by Berkeley's Center for Responsible, Decentralized Intelligence covering the full technological stack of agentic AI including foundation models, agent frameworks, evaluation methods, infrastructure, and deployment. Explicitly addresses critical safety and security challenges with a goal to advance responsible and secure agentic AI systems. Expected 5,000+ in-person attendees plus global livestream.

{
  "type_weight": 1,
  "source_trust": 0.85,
  "topic_relevance": 0.8,
  "time_proximity": 0.6327044025157232,
  "community_signal": 0.75,
  "speaker_org_signal": 0.85,
  "is_cfp_open": 1,
  "source_count": 1
}

5-day multi-track unconference bringing together researchers in theoretical AI alignment. Covers mathematical approaches including Singular Learning Theory, Agent Foundations, Causal Incentives, Computational Mechanics, and Scalable Oversight. Unconference format where participants can propose and lead sessions. Free to attend. Application deadline June 1. Limited travel and accommodation funding available needs-based. Limited onsite bedrooms available for booking after registration.

{
  "type_weight": 1,
  "source_trust": 0.85,
  "topic_relevance": 0.9,
  "time_proximity": 0.6226415094339622,
  "community_signal": 0.8,
  "speaker_org_signal": 0.8,
  "is_cfp_open": 1,
  "source_count": 1
}

Fully funded 5-month in-person AI safety fellowship pairing senior advisors with emerging talent on technical, governance, strategy, and field-building projects. Monthly stipend of $8,400, ~$15k compute budget for empirical fellows, workspace at Constellation Berkeley. Applications close Sept 26, onboarding completes Dec 31, program runs Jan 5 - Mar 31 2027 with extension through June 31. Over 80% of first cohort now working full-time in AI safety.

{
  "type_weight": 0.8,
  "source_trust": 0.85,
  "topic_relevance": 0.9,
  "time_proximity": 0.41132075471698115,
  "community_signal": 0.8,
  "speaker_org_signal": 0.85,
  "is_cfp_open": 1,
  "source_count": 2
}

OpenAI's safety fellowship program (Sept 2026 - Feb 2027) for researchers pursuing work on safety and alignment of advanced AI systems. Priority areas include safety evaluation, ethics, robustness, scalable mitigations, privacy-preserving safety methods, agentic oversight, and high-severity misuse domains. Monthly stipend, compute support, mentorship, and workspace in Berkeley. Application deadline May 3, notification July 25.

{
  "type_weight": 0.8,
  "source_trust": 0.9,
  "topic_relevance": 0.95,
  "time_proximity": 0.41132075471698115,
  "community_signal": 0.85,
  "speaker_org_signal": 0.95,
  "is_cfp_open": 1,
  "source_count": 2
}

UNIDIR's global conference bringing together diplomats, policymakers, military experts, industry leaders, academia, civil society, and research labs to discuss AI security and ethics. Part of UNIDIR's broader work on international AI policy, disarmament, and governance - directly relevant to the AI safety community's governance interests.

{
  "type_weight": 1,
  "source_trust": 0.85,
  "topic_relevance": 0.75,
  "time_proximity": 0.8540880503144654,
  "community_signal": 0.6,
  "speaker_org_signal": 0.8,
  "is_cfp_open": 1,
  "source_count": 1
}

Quarterly AI safety research fellowship based in London. Fellows pursue independent research projects with mentorship from alignment researchers. Applications for Q3 2026 cohort due May 3.

{
  "type_weight": 0.8,
  "source_trust": 0.75,
  "topic_relevance": 0.85,
  "time_proximity": 0.7987421383647799,
  "community_signal": 0.65,
  "speaker_org_signal": 0.7,
  "is_cfp_open": 1,
  "source_count": 1
}

ARENA's 8th cohort is a 4-5 week in-person AI safety bootcamp at LISA in Shoreditch, London. Covers technical skills for alignment research engineering. Provides travel, visa expenses, accommodation, and meals for all participants. Applications are now closed.

{
  "type_weight": 0.8,
  "source_trust": 0.9,
  "topic_relevance": 0.9,
  "time_proximity": 0.9748427672955975,
  "community_signal": 0.85,
  "speaker_org_signal": 0.85,
  "is_cfp_open": 0,
  "source_count": 2
}

12-week ML Alignment & Theory Scholars fellowship program connecting talented researchers with top mentors in AI alignment, interpretability, and governance. Research areas include AI control, scalable oversight, model organisms, model internals, model welfare, security, policy at AI compute/geopolitics/cybersecurity intersection, and agent foundations. $15k stipend, $12k compute budget, housing, meals, travel, research manager. Applications closed. ~75% continue in 6-12 month extension. Rated 9.4/10 average satisfaction.

{
  "type_weight": 0.8,
  "source_trust": 0.95,
  "topic_relevance": 0.95,
  "time_proximity": 0.939622641509434,
  "community_signal": 0.9,
  "speaker_org_signal": 0.95,
  "is_cfp_open": 0,
  "source_count": 1
}

Neural Information Processing Systems 2026 held across three satellite locations: Sydney, Atlanta, and Paris. Features Evaluations & Datasets Track, workshops, competitions, and safety-related tracks. Abstract deadline May 4, full paper deadline May 6, author notifications Sept 24. In-scope for safety-related workshops and eval track submissions.

{
  "type_weight": 1,
  "source_trust": 0.85,
  "topic_relevance": 0.75,
  "time_proximity": 0.16675675675675677,
  "community_signal": 0.8,
  "speaker_org_signal": 0.75,
  "is_cfp_open": 1,
  "source_count": 1
}

Technical AI Safety conference at Oxford Martin School. Free admission. Third iteration, first time in UK (previous in 2024 and 2025). Welcomes researchers and professionals from all backgrounds interested in AI safety discussions regardless of prior research experience. Organized by Oxford Martin AI Governance Initiative and Noeon Research. Registration now open.

{
  "type_weight": 1,
  "source_trust": 0.9,
  "topic_relevance": 0.95,
  "time_proximity": 0.8285714285714285,
  "community_signal": 0.85,
  "speaker_org_signal": 0.85,
  "is_cfp_open": 0,
  "source_count": 1
}

Foresight Institute's flagship conference bringing together leading scientists, entrepreneurs, funders, and policymakers to explore frontiers of science and technology. Features tracks on AI safety, security, and other emerging technologies. Part of Foresight's 40-year mission advancing transformative technology.

{
  "type_weight": 1,
  "source_trust": 0.75,
  "topic_relevance": 0.65,
  "time_proximity": 0.919496855345912,
  "community_signal": 0.5,
  "speaker_org_signal": 0.6,
  "is_cfp_open": 1,
  "source_count": 1
}

One-day alignment workshop bringing together global leaders in academia and industry to deepen collective understanding of AGI risks and explore mitigation strategies. Part of FAR.AI's global Alignment Workshop series following events in London, San Diego, Singapore, Vienna, New Orleans, and San Francisco.

{
  "type_weight": 0.85,
  "source_trust": 0.9,
  "topic_relevance": 0.95,
  "time_proximity": 0.7635220125786164,
  "community_signal": 0.8,
  "speaker_org_signal": 0.9,
  "is_cfp_open": 0,
  "source_count": 1
}

The inaugural International Programme on AI Evaluation is a 150-hour academic programme combining online lectures, hands-on courses, and an in-person capstone week in Valencia. With 40 fully-funded scholarships, it addresses the critical shortage of experts in AI evaluation needed by AI Safety Institutes and leading labs worldwide. Participants receive a 15 ECTS Expert Diploma.

{
  "type_weight": 0.8,
  "source_trust": 0.85,
  "topic_relevance": 0.85,
  "time_proximity": 0,
  "community_signal": 0.6,
  "speaker_org_signal": 0.85,
  "is_cfp_open": 1,
  "source_count": 1
}

3-month fellowship to launch or accelerate impactful careers in American AI governance and policy. Fellows conduct independent research projects under expert mentorship while building professional networks and developing policy expertise. Focus areas include public policy, political science, engineering, economics, biosecurity, cybersecurity, China studies, and risk management. Prioritizes bipartisan engagement, rigorous analysis, and practical policy relevance. $21,000 stipend plus travel support, weekday lunches, and DC office space. US work authorization required.

{
  "type_weight": 0.8,
  "source_trust": 0.9,
  "topic_relevance": 0.9,
  "time_proximity": 0.89937106918239,
  "community_signal": 0.8,
  "speaker_org_signal": 0.85,
  "is_cfp_open": 0,
  "source_count": 1
}

3-month fellowship for conducting independent research on AI governance topics. Fellows receive mentorship from field experts, participate in seminars and Q&A sessions, and build professional networks. Research outputs may include reports, white papers, journal articles, op-eds, or blog posts. £12,000 stipend plus travel support and weekday lunches. Open to candidates from government, academia, industry, or civil society with expertise in policy, political science, computer science, economics, or risk management. Visa sponsorship available.

{
  "type_weight": 0.8,
  "source_trust": 0.9,
  "topic_relevance": 0.9,
  "time_proximity": 0.89937106918239,
  "community_signal": 0.8,
  "speaker_org_signal": 0.85,
  "is_cfp_open": 0,
  "source_count": 1
}

Cambridge Boston Alignment Initiative's intensive nine-week summer fellowship for AI safety and biosecurity research. Fellows work on technical AI safety projects with expert mentorship in Cambridge, MA. Fully funded program covering housing, stipend, and research support.

{
  "type_weight": 0.8,
  "source_trust": 0.85,
  "topic_relevance": 0.85,
  "time_proximity": 0.9044025157232705,
  "community_signal": 0.7,
  "speaker_org_signal": 0.8,
  "is_cfp_open": 0,
  "source_count": 2
}

Anthropic's four-month fellowship program for AI safety research. Weekly stipend of $3,850 USD / £2,310 GBP / $4,300 CAD, plus ~$15k/month compute budget and close mentorship from Anthropic researchers. Priority areas include scalable oversight, adversarial robustness, and interpretability.

{
  "type_weight": 0.8,
  "source_trust": 0.95,
  "topic_relevance": 0.95,
  "time_proximity": 0.4571428571428572,
  "community_signal": 0.85,
  "speaker_org_signal": 0.95,
  "is_cfp_open": 0,
  "source_count": 1
}

Three-month fully-funded research fellowship for scholars in economics, law, international relations, and related fields focusing on societal impacts of advanced AI and institutions/policies for effective response. Fellows receive $25,000 stipend, covered travel, daily meals, and access to CAIS expertise and Bay Area network. Emphasizes producing publicly shareable research on AI's impact on economic distribution, corporate accountability, and geopolitical competition.

{
  "type_weight": 0.8,
  "source_trust": 0.9,
  "topic_relevance": 0.85,
  "time_proximity": 0.9345911949685535,
  "community_signal": 0.75,
  "speaker_org_signal": 0.9,
  "is_cfp_open": 0,
  "source_count": 1
}

Workshop on secure and verifiable AI development, bringing together researchers, builders, and funders across ML, hardware security, systems, cryptography, and computer security. Focuses on verification techniques for AI safety. Colocated with IEEE Security and Privacy conference. Organized by FAR.AI.

{
  "type_weight": 0.85,
  "source_trust": 0.9,
  "topic_relevance": 0.85,
  "time_proximity": 0.9428571428571428,
  "community_signal": 0.65,
  "speaker_org_signal": 0.85,
  "is_cfp_open": 0,
  "source_count": 1
}

Second Workshop on Technical AI Governance Research at ICML 2026, focusing on technical approaches to AI governance, policy, and regulation. Part of the main conference workshop track.

{
  "type_weight": 0.85,
  "source_trust": 0.85,
  "topic_relevance": 0.9,
  "time_proximity": 0.7383647798742139,
  "community_signal": 0.75,
  "speaker_org_signal": 0.7,
  "is_cfp_open": 0,
  "source_count": 1
}

Workshop at ICML 2026 focused on identifying, diagnosing, and fixing failure modes in agentic AI systems. Covers reproducible triggers for failures, diagnostic tracing methods, and verified repair approaches. Highly relevant to AI safety and robustness.

{
  "type_weight": 0.85,
  "source_trust": 0.85,
  "topic_relevance": 0.9,
  "time_proximity": 0.7383647798742139,
  "community_signal": 0.75,
  "speaker_org_signal": 0.7,
  "is_cfp_open": 0,
  "source_count": 1
}

Second Workshop on Agents in the Wild focusing on safety and security of AI agents deployed in real-world environments. Addresses challenges in ensuring safe and secure operation of autonomous agents. Part of ICML 2026 workshop track.

{
  "type_weight": 0.85,
  "source_trust": 0.85,
  "topic_relevance": 0.9,
  "time_proximity": 0.7333333333333334,
  "community_signal": 0.75,
  "speaker_org_signal": 0.7,
  "is_cfp_open": 0,
  "source_count": 1
}

Weekend AI safety hackathon focused on Global South participation and perspectives. Hybrid format allowing both online and in-person participation. Organized by Apart Research as part of their 55+ research sprints series with 6,000+ participants across 200+ global locations.

{
  "type_weight": 0.65,
  "source_trust": 0.85,
  "topic_relevance": 0.8,
  "time_proximity": 0.8490566037735849,
  "community_signal": 0.7,
  "speaker_org_signal": 0.7,
  "is_cfp_open": 0,
  "source_count": 1
}

Flagship conference gathering leading scientists, entrepreneurs, funders, and policymakers to explore frontiers of science and technology including AI safety. Foresight is a 40-year-old organization focused on transformative technology. Registration available.

{
  "type_weight": 1,
  "source_trust": 0.75,
  "topic_relevance": 0.65,
  "time_proximity": 0.919496855345912,
  "community_signal": 0.6,
  "speaker_org_signal": 0.65,
  "is_cfp_open": 0,
  "source_count": 1
}

Workshop at ACL 2026 focused on AI evaluation in practice, centering tensions and collaborations between model developers and evaluators as generative AI systems are increasingly integrated into real-world products and decision-making pipelines. Organized by EvalEval Coalition.

{
  "type_weight": 0.85,
  "source_trust": 0.75,
  "topic_relevance": 0.8,
  "time_proximity": 0.6276729559748427,
  "community_signal": 0.6,
  "speaker_org_signal": 0.65,
  "is_cfp_open": 0,
  "source_count": 1
}

Two-day workshop on AI safety and security in Berlin. Focus on secure and sovereign AI systems. Organized by Foresight Institute. Registration available.

{
  "type_weight": 0.85,
  "source_trust": 0.75,
  "topic_relevance": 0.75,
  "time_proximity": 0.7031446540880504,
  "community_signal": 0.6,
  "speaker_org_signal": 0.65,
  "is_cfp_open": 0,
  "source_count": 1
}

9th annual Cognitive Computational Neuroscience conference at NYU. Forum for discussion among researchers in cognitive science, neuroscience, and AI on computations underlying complex behavior. In-scope for AAE due to neural network interpretability, predictive coding, and metacognition measurement sessions. Single-track format with keynotes from Brenden Lake, Ila Fiete, Kenji Doya, Doris Tsao, and Alona Fyshe.

{
  "type_weight": 1,
  "source_trust": 0.8,
  "topic_relevance": 0.7,
  "time_proximity": 0.6226415094339622,
  "community_signal": 0.55,
  "speaker_org_signal": 0.7,
  "is_cfp_open": 0,
  "source_count": 1
}

Two-day conference on cybersecurity and technology policy covering AI security topics. Part of UNIDIR's Security and Technology Programme addressing implications of AI for international peace and security. In-scope for AAE tracker as international AI policy/disarmament falls within governance community's concerns.

{
  "type_weight": 1,
  "source_trust": 0.85,
  "topic_relevance": 0.65,
  "time_proximity": 0.5428571428571429,
  "community_signal": 0.5,
  "speaker_org_signal": 0.6,
  "is_cfp_open": 0,
  "source_count": 1
}

Two-week intensive summer school at Columbia University covering machine learning topics including mechanistic interpretability, alignment/safety, RAG & agents, and LLM systems. Approximately 200 PhD students participate alongside faculty and industry speakers. In-scope due to dedicated alignment and mechanistic interpretability tracks.

{
  "type_weight": 0.35,
  "source_trust": 0.75,
  "topic_relevance": 0.7,
  "time_proximity": 0.869182389937107,
  "community_signal": 0.5,
  "speaker_org_signal": 0.6,
  "is_cfp_open": 0,
  "source_count": 1
}

A free, accessible workshop hosted by AI Safety Awareness Group Oakland exploring AI's trajectory and societal impact. No technical background required. Features live demonstrations of current AI systems, interactive forecasting activities, and discussions about AI's implications for work, relationships, and society over the next 1-5 years.

{
  "type_weight": 0.45,
  "source_trust": 0.7,
  "topic_relevance": 0.7,
  "time_proximity": 0.5714285714285714,
  "community_signal": 0.6,
  "speaker_org_signal": 0.5,
  "is_cfp_open": 0,
  "source_count": 1
}