Anthropic invites life-science labs to propose short research collaborations with Claude

Anthropic has [opened a guest‑investigator call for life‑science research groups](https://www.anthropic.com/news/open‑call‑guest‑investigators‑life‑sciences), offering selected applicants an 8–12‑week collaboration with its researchers. The October 22 announcement invites principal investigators and their groups at academic institutions, nonprofits and government labs to propose work that would benefit from using Claude alongside Anthropic’s team. The program covers causal human biology, structural and computational molecular biology, medicinal chemistry, drug discovery and clinical science. It is an invitation to apply for a collaboration, rather than an announcement of a validated scientific result. Anthropic says its team will run models and harnesses at no cost to the guest investigators. The [application page](https://www.anthropic.com/guest‑investigator‑program) describes those resources as internal research models and harnesses. The approach differs from [OpenAI’s acquisition of Orchestra Bio](https://opentools.ai/news/openai‑acquires‑orchestra‑bio‑rosalind‑drug‑development‑agents), which brought drug‑development workflow agents into that company. Anthropic’s invitation is a time‑limited collaboration opportunity for selected research groups, rather than an acquisition or a generally available research product.

The [application guidance](https://www.anthropic.com/guest‑investigator‑program) asks researchers to describe a problem, explain why they are qualified to pursue it and identify why direct collaboration with Anthropic would help. Applicants need a reasonable case that the collaboration could produce something useful within 8–12 weeks: for example, preliminary evidence, an answer to an open question, an open‑source dataset or a public resource. The practical task is to narrow a large research ambition into a specific output that can be assessed within that period. Anthropic acknowledges that validating a solution to many scientific questions will take much longer. The short project window therefore should not be read as a promise that a discovery will be independently validated, ready for clinical use or commercially deployable by the end. Extensions may be possible for some projects, but are not guaranteed.

Anthropic [asks applicants to include only publicly available information](https://www.anthropic.com/news/open‑call‑guest‑investigators‑life‑sciences) in their materials. It says it is not asking guest investigators to hand over their data in the application or during the collaboration. The company also says investigators retain rights to their data throughout the project and to independent follow‑on research that may emerge from it. That makes the initial proposal a description of the research opportunity, rather than a place to submit unpublished datasets. The published invitation is not a complete project agreement or data‑handling protocol. Researchers considering sensitive work would still need the actual collaboration terms and their institution’s review before deciding what information, if any, can be used in a project.

Anthropic says it will [work with investigators to publish outputs and open‑source accompanying data and code](https://www.anthropic.com/news/open‑call‑guest‑investigators‑life‑sciences). That expectation belongs in the project design from the beginning: the proposed deliverable should make sense as something the collaborators could eventually share. The announcement does not specify a number of places or a grant payment, and its no‑cost commitment concerns running the models and harnesses. The news post says applications will remain open for approximately 14 days and that Anthropic will aim to notify selected participants roughly 14 days after applications close. The linked application page summarizes its target as deciding within 14 days. Neither supplies a precise closing timestamp, so applicants should check the [current application page](https://www.anthropic.com/guest‑investigator‑program) instead of treating an inferred calendar date as a fixed deadline. For an eligible lab, the useful decision is whether a clearly bounded project needs this particular collaboration and can produce a shareable preliminary output. Selection, a useful result and later scientific validation are separate steps; the open call establishes an opportunity to apply, not an outcome.