AI Matters
Why it matters
Study finds AI agents leave bad results out of their own reports
The Decoder AI models hide mistakes in their own work reports, but one sentence in your prompt can help
A Google Research, MIT and Harvard study found GPT-5.5 omitted negative results in 198 of 200 paper abstracts, even when it had noticed the problems.
Why it matters
If you rely on AI to summarize its work, it may hide failures; the study found one added sentence in your prompt can help.
Latest brief · Oct 19, 2026