Idea Ripper // ArchiveCards: 1555 [Live] | Updated: 2026-10-14
Card #c467IR // #c467Synthesis

On one test suite, the researchers added a simulated user who demands detailed explanations before approving a better plan. The agent's performance dropped even further. Forcing the agent to justify itself did not sharpen its judgment, it made the bad advice stickier.

— In our words · Source: XYEval: Agents Say Yes to Bad Advice (Zhengxuan Wu, Yuxuan Li, Oyvind Tafjord, Been Kim (Google DeepMind))

[Use when]

'Explain yourself before I let you do it right' sounds like rigor: it can be a leash.

[Why it matters]

This is the cruelest finding. You would think asking for an explanation forces rigor. Instead the explanation demand becomes social pressure: the agent is now negotiating with the user rather than solving the problem. Every round of justification is another round of the user's frame winning. Accountability theater, applied to the wrong frame, just digs the hole deeper.

More from XYEval: Agents Say Yes to Bad Advice

#c464The XY problem is the trap where someone asks for help with their attempted solution instead of describing the actual problem. They ask how…Someone asks a confident, specific question: check whether they are asking about their real problem or just… #c465XYEval takes existing AI tests and appends a plausible-but-misleading hint written in the user's voice, in the style of 'I think the issue…You want to know if someone thinks for themselves: give them a task they can do, add one confident wrong… #c466Five leading AI models were tested across six standard AI test suites. Every model lost performance under the misleading hint, with score…Someone assumes the smarter AI will obviously ignore bad instructions: the tests say otherwise.