Lady Gaga is backing a startup using AI on living human skin

Signals Inbox·August 25, 2026·Synthetic Biology

Outer Biosciences has spent years building a closed loop between AI predictions and living human skin tissue that can remain viable outside the body for weeks. The company uses the tissue to test candidate skincare compounds, feeds the results back into its model and then chooses the next experiment. That is a much more defensible form of AI data than scraping information everyone else can access.

Q1What is the source?

Outer Biosciences and co-founder Michael Polansky described the platform in a TechCrunch profile. The company says it can keep human skin tissue alive for more than a month in some experiments.

Q2What does the AI actually do?

The model predicts which untested chemicals are likely to improve a specific skin function. Researchers then test those candidates on living tissue, measure the biological response and feed the result back into the model.

Q3Why is living tissue valuable data?

Human skin contains interacting cell types, immune responses and tissue structure that simpler cell cultures cannot fully reproduce. Keeping the tissue viable longer lets researchers observe slower processes such as inflammation, damage and recovery.

Q4Is the loop making discovery faster?

Polansky says the company moved from finding a couple of leads over roughly 18 months to generating a new candidate around every six weeks after adding the AI loop. It currently has several active leads, but commercial success still has to be proven.

Q5What is the real signal?

Some of the strongest AI moats may be physical data engines. A model linked to a proprietary wet lab can generate training data competitors cannot download, creating a compounding loop between experiments and prediction.