Organisation Name:
Anthropic
Application Deadline:
September 21, 2026
About Organisation
Anthropic is an AI research and technology company focused on developing reliable, interpretable and steerable artificial intelligence systems. Its products include Claude, an AI assistant, while its research spans areas such as AI safety, alignment, reinforcement learning, interpretability and model behaviour. Through this new grant programme, Anthropic is supporting independent researchers working to better understand and measure how AI systems affect users’ wellbeing. The company aims to encourage rigorous, open-source evaluations that can help researchers and developers establish stronger standards for responsible AI behaviour.
About the Grant
The Anthropic Wellbeing Research Grants programme is a US$5 million funding initiative supporting independent research into AI’s effects on user wellbeing. The programme is designed to improve the quality of evaluations and benchmarks used to assess AI behaviour, particularly in sensitive or evolving conversations. Selected researchers will receive direct funding, access to Anthropic models and technical support while retaining independence over their work. Projects are expected to produce open-source evaluations that can be used by developers and researchers across the AI ecosystem. Applications are due September 21, 2026, with selected applicants expected to be invited to submit full proposals.
Who can Apply?
The programme is aimed at independent researchers and subject-matter experts working on AI wellbeing evaluation and benchmarking. Potential applicants include clinicians, psychologists, methodologists and researchers with relevant expertise. Strong proposals should clearly define what the evaluation measures and how success or failure is determined. Anthropic particularly seeks projects involving clinical or subject-matter experts in design and validation, assessing both harmful behaviour and excessive refusal, modelling realistic multi-turn conversations, and validating automated graders against real experts. The programme is international in scope based on Anthropic’s description, although applicants should review the application form for any specific submission requirements. Anthropic
How to Apply?
Applicants should review Anthropic’s official guidance on developing rigorous wellbeing evaluations and complete the online application form before the deadline. Proposed research should clearly explain the wellbeing outcome being evaluated, methodology, validation approach and relevance to real-world AI use. Strong applications should account for changing conversational context and evaluate both overcompliance and overrefusal. Anthropic states that applicants selected for the next stage will be notified by October 5, 2026.
Application Deadline
The application deadline is September 21, 2026
Last Date To Apply
For more information, please visit the grant website here.
