New NASA A.I. Effort For Mars Exploration Ignores The Search For Life (e.g. Astrobiology)

Keith’s note: In case you missed it the White House recently went in – big time – on global AI leadership – here’s the plan at ai.gov. One would assume that NASA was paying attention. They did to some extent. NASA SMD just issued C.12 Foundational Artificial Intelligence for the Moon and Mars (FAIMM) stating that it is “Amended January 13, 2026: This amendment presents this new program element in ROSES-2025.” One of the prime reasons to explore Mars in the first place for the past six decades with robotics and humans has been the search for life – past or present. The NASA program for this is called “Astrobiology”. Yet no mention of the words “astrobiology” or “exobiology” or “life” or “biology” is made in c.12. There is no mention on the new NASA Astrobiology site either. Why is that? More below.

There is a big push for the use of Artificial Intelligence(AI) at NASA Astrobiology. A recent paper titled “Foundation Models for Astrobiology: Paper I — Workshop and Overview“ about a NASA workshop is dripping with references to AI and Machine Learning. There is an AI-Astrobiology community hosted at NASA Ames but it makes no mention of this new effort outlined in NASA ROSES-2025 C.12

It is baffling that one part of NASA pushes for AI in Astrobiology yet the other parts of NASA that do AI or Astrobiology ignore c.12 – and each other. And given that the whole Mars exploration effort has life science its nexus point makes this even more baffling.

[FYI you can find more AI/Astrobiology postings here at Astrobiology.com]

Here is what the NASA ROSES-2025 C.12 announcement says about what it encompasses. Again, given the long standing emphasis on the whole life-on-Mars issue you’d expect something descriptive of that aspect of the AI efforts to be mentioned in that context. Again, no mention of Astrobiology is made.

C.12 Foundational Artificial Intelligence for the Moon and Mars:

Foundational Artificial Intelligence for the Moon and Mars (FAIMM) will enable individual researchers to participate as members of existing teams of scientists and artificial intelligence (AI) researchers who are designing science and exploration applications for large, general AI models known as Foundation Models (FMs). FMs harness large datasets to, e.g., transform science and exploration on the Moon and Mars and can be applied to a range of AI and Machine Learning (ML) tasks. Example applications include crater detection, feature identification, landing site assessment, and assessing evidence for the presence of water ice. The Science Mission Directorate (SMD) Office of the Chief Science Data Officer is supporting the development of FMs for each division in the Science Mission Directorate by the NASA Interagency Implementation and Advanced Concepts Team (IMPACT), and this program element aims to maximize the benefit of this effort to planetary science and exploration.

In this collaborative and interdisciplinary effort, selected participants will work with each other, existing project team members, and AI researchers and engineers on one of two teams, one team working on applications for a Lunar FM and one team working on applications for a Martian FM. This program seeks to expand the personnel, team skills and expertise, datasets, and science and exploration disciplines contributing to large AI models, and no prior AI/ML experience is required. The FM, training data, and source code will be made publicly available as an open source, open weight model. Open weight refers to publishing the model parameters derived during training.

Each team (Moon or Mars) will collaborate on the following objectives:

1. Pilot the use of an AI FM to science and/or exploration applications for the target
   body.
2. Identify and/or develop benchmarks for transparent, reproducible, and rigorous
   assessment of the FM’s accuracy.
3. Contribute to the assessment of sources of error in the FM and assess their
   impact on the applications.
4. Contribute to open development of the code base and applications.
5. Create documentation on how to develop applications, including the process for
   fine tuning the FM to the application and lessons learned, for open public
   distribution to enable AI adoption.
6. Create documentation on how the applications are used, including example code
   and datasets, for open public distribution to enable AI adoption.
7. Contribute to publication of science results – including uncertainties and
   assumptions — in peer reviewed journals.

NASA email announcement issued on 13 January 2026:

C.12 Foundational Artificial Intelligence for the Moon and Mars (FAIMM) is intended to enable individual researchers to participate as members of teams who are designing science and exploration applications for large, general artificial intelligence (AI) models known as Foundation Models (FMs) for the Moon and Mars. These FMs harness large datasets to transform science and exploration on the Moon and Mars and can be applied to a range of AI and Machine Learning (ML) tasks. In this collaborative and interdisciplinary effort, selected participants will work with each other, existing project team members, and AI researchers and engineers. This program seeks to expand the personnel, skills and expertise, datasets, and science and exploration disciplines contributing to large AI models, and no prior AI/ML experience is required.

ROSES-2025 Amendment 37 presents C.12 FAIMM as a new program element in ROSES-2025. Neither Step-1 proposals nor NOIs are requested for this program element. Proposals are due by April 28, 2026.

On or about January 13, 2026, this Amendment to the NASA Research Announcement “Research Opportunities in Space and Earth Sciences (ROSES) 2025” (NNH25ZDA001N) will be posted on the NASA research opportunity homepage at https://solicitation.nasaprs.com/ROSES2025 and will appear on SARA’s ROSES blog at: https://science.nasa.gov/researchers/solicitations/roses-2025/

Questions concerning C.12 FAIMM may be directed to Rebekah Dawson-Rigas at HQ-FAIMM -at – mail.nasa.gov