Resources

Alongside our Haas AI Research Living Lab grant program, you may be interested in or already exploring external funding opportunities offered to academic researchers, such as:

• Anthropic AI for Science
• Anthropic Claude Team plan for scientists
• OpenAI Researcher Access Program
• Google Gemini Academic Program
• Google Cloud research credits
• Anthropic Economic Futures Research Fund

While you are welcome to explore these opportunities, please coordinate with our Berkeley Intellectual Property & Industry Research Alliances (IPIRA) on your proposal. IPIRA is currently leading a long term process to negotiate acceptable terms with these funders. Please have a look at this Corporate Research Awards page for these types of research awards.

Check in with Haas Staff AI Resources for currently available AI tools within our Haas & Berkeley infrastructure.

If one of your projects/proposals works with tools that are currently not available here at Berkeley, reach out and let’s see how to work through this together!

In August 2026, the Office of Research Integrity (ORI) in the U.S. Department of Health and Human Services (HHS) issued new federal guidance on generative AI as it relates to the Research Misconduct Policy.

UC Berkeley follows the ORI Research Misconduct Policy, which applies to all research. The new federal guidance states that research misconduct standards continue to apply when generative AI is involved, and researchers remain responsible for all aspects of their work. The new guidance establishes the need for disclosing the use of generative AI, as well as the importance of verifying all AI-assisted and/or AI-derived results for accuracy and validity. Key points from the New Federal Guidance on Generative AI:

• RESEARCH MISCONDUCT STANDARDS continue to apply when generative AI is involved. As a reminder, research misconduct is defined as:
  • Fabrication, falsification, or plagiarism in proposing, performing, or reviewing research, or in reporting research results;
  • It does not include honest error or differences in opinion;
  • A finding of research misconduct requires a significant departure from accepted practices, conduct that is intentional, knowing, or reckless, and proof by a preponderance of the evidence.

• DISCLOSURE: Researchers should disclose the use of generative AI in research reports, manuscripts, and publications. For grant proposals, researchers should follow the agency’s disclosure requirements. Disclosure may support reproducibility and help address misconduct allegations. However, it does not preclude a finding of research misconduct. Disclosure samples are available here. Disclosure should include:
  • Tool Identification: Specific platform, model version, and access date
  • Location/Scope: Exact sections, figures, or code files where AI was applied.
  • Purpose: Specific tasks performed by generative AI.
  • Human Verification: Explicit assertion that humans verified facts, citations, and data to prevent Fabrication, Falsification, and Plagiarism.
  • Data Protection Notice: Confirmation that no protected health information (PHI/HIPAA), unpublished data, or trade secrets were exposed to non-approved external AI models.

• REVIEW & VERIFY: Researchers are responsible for all aspects of their research and should review and verify their AI-assisted work:
  • To ensure accuracy and validity of research data and results
  • To confirm appropriate citation of sources and correct references

Information on this new guidance can be found on the VCRO’s webpage on Responding to Allegations of Research Misconduct.  If you have questions about research misconduct, please contact Research Integrity Officer and Associate Vice Chancellor Lok Siu at [email protected].

Check in with the Berkeley AI Hub for most up-to-date articles and news.

• Berkeley Artificial Intelligence Research (BAIR) Responsible AI Initiative led by Haas Professional Faculty Genevieve Smith has several projects and events for your consideration.
• Berkeley Professor Ted Miguel is a member of OpenAI’s Economic Research Exchange working on high-impact external research on the economic effects of AI.

There are calls to ‘pace the frontier’ in AI. Berkeley resources to consider when thinking through the risks related to AI that can inform research agendas include:

• Center for Human-Compatible Artificial Intelligence (CHAI) – A leader in artificial intelligence research (see list of publications here), Berkeley Professor Stuart Russell recently wrote and participated in several discussions (see here and here and here).
  • See Berkeley alum’s contribution to the Hugging Face investigation.

• CITRIS and the Banatao Institute – ED Camille Crittenden serves on the UC AI council. See recent interview on California State legislation toward improving AI safety and reducing catastrophic risk (AB1405 and SB813).