Schneier Flags New Research: Humans Play More Cautiously Against AI Than Against Other Humans
What Happened
Bruce Schneier highlighted a new study showing participants shifted strategy meaningfully when told their counterpart was an AI, selecting more conservative moves and citing both the model's reasoning power and an unexpected perception of cooperative intent. The effect was strongest in participants with high strategic-reasoning scores.
My Take
This is a soft but important finding for anyone deploying AI in negotiation, pricing, or customer-service contexts: disclosing that the counterparty is AI changes human behavior in measurable and monetizable ways. Enterprises that hide AI behind a human persona are leaving value on the table in some contexts and creating legal exposure in others. The smart move for 2026 is to A/B test disclosed-vs-undisclosed AI interaction across your funnel, because "does the customer know" is about to become a regulated variable under the EU AI Act's transparency code due in May–June.
Read Original Source