# Inherent says Faraday replicated papers ahead of larger AI models

_Saturday, August 22, 2026 at 3:00 PM EDT · AI · Latest · Tier 2 — Notable_

![Inherent says Faraday replicated papers ahead of larger AI models — Primary](https://techcrunch.com/wp-content/uploads/2026/08/inherent-labs-cofounders-credit-Anna-Gordon.jpeg?resize=1200,800)

London AI lab Inherent says its newly released Faraday agent independently reproduced findings from published scientific papers without being given the answers, outperforming Anthropic's Claude Opus 4.8 and OpenAI's GPT-5.5 on that task. The company says Faraday runs on Qwen 3.6, a 27-billion-parameter model, and uses reinforcement learning to train experimental choices; it also uses OpenAI's GPT-5.5 Codex as a coding tool. Inherent emerged from stealth weeks ago with a $50 million seed round.

## Sources

- [TechCrunch](https://techcrunch.com/2026/08/22/inherent-founded-by-deepmind-alumni-says-its-ai-teammate-just-outperformed-anthropic-and-openai-at-replicating-research/)

---
Canonical: https://techandbusiness.org/newswire/CyYTWICHWA-OvqJ14PbaEB
Retrieved: 2026-08-22T20:34:27.869Z
Publisher: Tech & Business (techandbusiness.org)