Open deadlines

Two-day interdisciplinary forum exploring technical AI safety challenges, governance approaches, risk assessment, evaluation standards, and cross-sector collaboration. Grounded in the International AI Safety Report. Features keynotes, panels, workshops, and networking. Organized by Gradient Institute with RAND, Timaeus, Good Ancestors, and CSIRO Data61. Builds on Australia's AI Safety Institute and National AI Plan.

UNIDIR flagship conference spotlighting how artificial intelligence is reshaping global security. Brings together experts to discuss AI's intersection with peace and security issues. Part of UNIDIR's Strategic Framework 2026-2030 emphasis on AI as primary focus area. Organization produces 132 publications yearly and reaches 193 states through research with 20,000+ annual event participants.

Regional AI safety competition focusing on tools, evaluations, and policy research from Latin America, Africa, or Asia. Sponsored by Schmidt Sciences with pipeline to fellowship and placement opportunities. Online and in-person format enabling global participation from the Global South.

10-week fellowship for researchers and entrepreneurs working on mitigating risks from frontier AI. Targets talented individuals at any career stage interested in AI safety and governance research. Provides competitive stipend, meals during working hours, and coverage for transport, visas, and lodging. Hosts 30+ events with mentorship from expert researchers.

3-week ML upskilling bootcamp for AI safety focusing on interpretability and RL, based on ARENA curriculum. Run by Cambridge Boston Alignment Initiative. Provides housing, meals, 24/7 office access, dedicated teaching assistants, and travel support to participants.

Four-month full-time paid AI safety research fellowship providing funding, compute (~$15k/month), and direct mentorship from Anthropic researchers to work on real safety and security projects. Focus areas include scalable oversight, adversarial robustness, and other AI safety topics. Over 40% of fellows from first cohort joined Anthropic full-time.

Workshop on Evaluating Evaluations (EvalEval) at ACL 2026. Focuses on evaluation methodology, measurement theory, validity, reliability, reproducibility, infrastructure costs, and sociotechnical impacts including bias and privacy. Features panel with model developers, oral presentations, posters, and Every Eval Ever shared task standardizing results across 22,000+ models.

Workshop bringing together global leaders in academia and industry to explore risk mitigation strategies for artificial general intelligence. Organized by FAR.AI. Co-located with ICML 2026 in Seoul.

AI safety conference in New Zealand organized by BlueDot Impact community. First major AI safety conference in New Zealand, bringing together researchers, practitioners, and policy experts in the region to discuss technical safety, governance, and alignment challenges.

Foresight Institute technical workshop bringing together researchers, engineers, and entrepreneurs in computer science, ML, crypto, security, and related fields. Focus on secure AI development, safety infrastructure, and sovereignty considerations for AI systems. Part of Foresight's Secure AI programme supporting frontier science approaches to AI safety.

Competition focused on improving estimation algorithms for random MLPs. Partnership between ARC (Alignment Research Center) and AIcrowd. Warm-up round focuses on developing theoretical foundations for mechanistic explanations of neural network behavior by analyzing weights directly rather than relying on sampling methods.

3-week intensive program in applied mathematics for AI alignment by Iliad. Based in Berkeley at Lighthaven. Provides $5,000 support. Selection based on estimated mathematical strength. Part of Iliad's series of intensive and fellowship programs.

3-week ML upskilling bootcamp for AI safety focusing on interpretability and RL, based on ARENA curriculum. Run by Cambridge Boston Alignment Initiative. Provides housing, meals, 24/7 office access, dedicated teaching assistants, and travel support to participants.

Regional EA Global conference with heavy AI safety attendance. Berkeley location draws alignment researchers, interpretability practitioners, and governance community. In scope for social and community networking reasons.

3-month mentored research fellowship in applied mathematics for AI alignment by Iliad. Based in London at LISA. Provides $6,000/month stipend. Selection based on estimated mathematical strength.

4-week intensive program in applied mathematics for AI alignment by Iliad. Based in London at LISA. Provides $5,000 support. Selection based on estimated mathematical strength.

Neural Information Processing Systems conference with three satellite locations. Includes dedicated Ethics Review track, Reproducibility as official track, and Responsible AI metadata requirements in Evaluations and Datasets Track. Features Code of Conduct and Code of Ethics fostering diverse, inclusive community. Safety-related workshops cross-referenced with keywords.

4-week intensive program in applied mathematics for AI alignment by Iliad. Based in Berkeley. Provides $5,000 support. Selection based on estimated mathematical strength.

EA Global conference with heavy AI safety attendance and programming. Annual gathering bringing together effective altruism community members working on global priorities including AI safety, alignment, and governance. In scope for social and community event reasons.

4-week intensive program in applied mathematics for AI alignment by Iliad. Based in London at LISA. Provides $5,000 support. Selection based on estimated mathematical strength.

Foresight Institute flagship Vision Weekend conference gathering leading scientists, entrepreneurs, funders, and policymakers to explore frontiers of science and technology. Features AI safety track alongside other frontier science themes. 40-year-old organization focused on transformative technology with established community presence.