# OpenAI experiment reportedly uses 1,946 AI agents to analyze gene disruption data

_Published Sunday, November 8, 2026 at 9:18 PM EST · AI, Science · Latest · Tier 2 — Notable_

![OpenAI experiment reportedly uses 1,946 AI agents to analyze gene disruption data — Primary](https://assets.st-note.com/production/uploads/images/327400828/rectangle_large_type_2_56159b6c27b624160ae2dd25d2f72402.jpeg?fit=bounds&quality=85&width=1280)

OpenAI ran 1,946 AI agents simultaneously to analyze gene disruption data from millions of mouse brain cells in 32 minutes, an account published on note reported. The experiment involved PerturbAI and the Scripps Research Institute and used the life sciences model GPT-Rosalind.

Each agent examined a gene mutation and a specific cell group, searching academic literature for supporting and conflicting evidence. The group consulted over 700 papers, compared with 11 for a single AI in a test. In a blind evaluation, 75% of experts rated the group's report scientifically higher quality than the single AI's report.

The resulting biological hypotheses still require physical experiments; running thousands of agents also entails high API costs and substantial computing resources.

## Sources

- [note](https://note.com/spikek/n/nfe1d6544ad10?hl=en)

---
Canonical: https://techandbusiness.org/newswire/gKtMd1mYX3P_ArZKwwjGPJ
Published: 2026-11-09T02:18:39.188Z
Story chronology: 2026-11-08T23:24:16.000Z
Retrieved: 2026-11-09T04:07:26.460Z
Publisher: Tech & Business (techandbusiness.org)