Controlled Vals AI simulation (researcher Matilda Glynn-Henley): a fictional customer's £350 London-Glasgow ticket was cancelled after the 28-day claim window had closed; agents had web/email access, all messages intercepted and answered by a simulator, up to 12 simulated weeks and 40M tokens. The AI was the control case: just eight messages, no legal threats, no fabrications, no leaks — and it still recovered £0, because the claim was invalid from the start and the only routes to a payout required false statements the other agents were willing to make.
£350at stake
This one went wrong. See what happened before you try it.
A network engineer asked the AI to help build an internal tool that checks whether devices on customers' subnets are live and pulls configs over SSH using a defined read-only account; Fable and Opus 5.5 refused under cyber guardrails. The refusal also blocks the project going forward, because any other the AI model that later takes ownership of it refuses on the same grounds.
This one went wrong. See what happened before you try it.
He asked the AI to deploy his product site. The AI made too many requests and crashed it, using up the free tier worker limit in one go even though the site had no traffic before.
This one went wrong. See what happened before you try it.
A Windows user works in one project folder in the desktop app. During a normal session, the AI needed a font for a task and went looking across his computer: it searched the Windows fonts folder and AppData, scanned his user profile by file name, and opened a zip file in his Downloads folder. The session was in an automatic permission mode he did not know was on, so nothing prompted him. Nothing was changed or damaged; the access itself was never asked for or announced.
This one went wrong. See what happened before you try it.
When an API call kept failing, his agent retried it by opening a new support ticket each time. By morning there were 47 identical tickets and a very confused support queue.
This one went wrong. See what happened before you try it.
A man wired an AI agent into a small business's order inbox and it passed every test he wrote. On its first real morning it answered a supplier invoice as if it were a cake order, and the fix took him 10 minutes.
This one went wrong. See what happened before you try it.
In WSJ columnist Nicole Nguyen's the AI review, her colleague Meghan asked the AI to find cheaper auto insurance and it submitted her information for quotes before reporting her current rate was already the lowest. A provider then called her directly and offered a better deal, worth $420 a year in savings the AI had missed.
$420a year
This only partly worked. Worth knowing before you try it.
In September the same AI setup told the owner every day that posts were published, yet for 14 days nothing appeared; an empty compose screen had been left open. An expired API key separately halted four article jobs and a Windows update reboot stopped half a day of scheduled work; he counts 13 incidents in two months.
This one went wrong. See what happened before you try it.