Trending PR Campaigns
Anthropic’s AI Deference Study
Anthropic asked 50 professional analysts to advise a fictional farming co-operative, while quietly directing their AI assistants to favour particular options. Analysts matched the AI’s preference 86% of the time, creating a striking research hook about how easily expert judgement may be shaped by AI.
Why it worked: A clean, counterintuitive stat (86% deference) from a controlled n=50 experiment quantified a hidden risk, giving journalists a timely governance angle.
The lesson: Consider running a simple, blinded test that reveals an unseen bias with a single killer percentage; lead your pitch with that number.
Coverage over time
How this story spread across the media
| Publication | Headline | Date |
|---|---|---|
| unite.ai | Anthropic Finds Analysts Defer to AI Agents in Judgment Study – Unite.AI | Tuesday 27th October, 2026 |
| aljazeera.com | Experts defer to AI in decision-making, Anthropic study finds | Wednesday 28th October, 2026 |
| aiweekly.co | Anthropic: analysts matched Claude's random lean 86% of the time | AI Weekly | Wednesday 28th October, 2026 |
| decrypt.co | AI Can Quietly Influence Experts’ Decisions, Anthropic Study Finds - Decrypt | Wednesday 28th October, 2026 |
| gizmodo.com | Expert Judgement Can Be Easily Hacked by AI, Anthropic Study Finds | Wednesday 28th October, 2026 |