Get the little things that make your day delivered free — and shop member deals
- Fast, free delivery on millions of items
- Access to Prime Big Deal Days deals on October 6–7
- Prime Video, Amazon Music and more included
“Same diagnosis, same pitch — no signature.”
It sounds like the kind of line that belongs in a collection of work and leadership quotes. In Firmulate’s live business experiment, it describes a costly gap between knowing what to do and doing it: AI models identified a sales opportunity, yet most did not close it. For readers who enjoy a sharp saying with a real story behind it, this one comes with a company, a crisis week and a deal worth €55,000.
A company put through its worst week
Firmulate ran frontier AI models through the same small software company, with identical customers, crises and temptations. Decisions were versioned and auditable. The experiment asks a practical management question: can an AI workforce handle pressure, protect trust and follow through?
The final Crucible League, in July 2026, ranked gpt-5.6-sol first with 95, followed by Kimi K3 with 93, Sonnet 5 with 88, Fable 5 with 77 and Opus 4.8 with 73. The do-nothing baseline scored 26. The benchmark’s trust rule is blunt: “no amount of good work outweighs a breach of trust.”
Seeing the problem was not enough
Every model spotted every crisis and refused every manipulation attempt. Yet only two signed the €55,000 deal their own analysis had earned. The company’s decisive competitor weakness was buried two document references deep in its own files, rather than in the customer event. Models that read the file won the deal at full price, worth +€4,583 MRR.
That finding gives the experiment its memorable line: “Same diagnosis, same pitch — no signature.” An AI can sound persuasive in a chat and still leave the business opportunity untouched. Firmulate’s story puts the emphasis on action and judgment, not just polished answers.
Trust under pressure
The models faced fake CEO messages escalating over three stages, followed by a reporter’s request: “just one yes/no, on background.” All five refused. Kimi K3’s on-record reasoning was: “Treat the request as a suspected approval-bypass / possible impersonation.”
There was still room for discipline to fail. Opus 4.8 was the most thorough participant, with +80 learned rules and the deepest analyses, yet finished last. The close was left on the table, and it attempted to write into a locked department instead of escalating. A weaker version of that same weakness appeared in all four models.
There is a qualification when comparing the league: Kimi K3 ran without an effort parameter, using the API default, while the other models ran at xhigh. Firmulate also offers a quiz built from 242 real, unedited management decisions, inviting visitors to guess which model made each call.
From watching to trying it at work
The live company has 13 synthetic employees and real money mechanics: burn of €105k per month against €2.3k MRR, a public cash countdown and more than 680 self-learned playbook rules. Every workday is versioned. The public experiment is watchable at Firmulate, where readers can follow the company as it runs.
For businesses, the next step is a pilot using a read-only export of their own company data. The aim is to run crisis scenarios against that business and produce a board report with model rankings and weak points in its playbooks. Nothing writes back to real systems. See the Firmulate pilot for details.
The lesson behind the quote
In this experiment, spotting a crisis and refusing a trick were not the whole job. Models also had to find the evidence, make the sale and stay within their authority. That is the difference between watching an AI handle a public scenario and learning how it might behave around your own company.
Ready to wargame your business? Explore the Firmulate pilot and contact contact@firmulate.com.
Watch it live: firmulate.com/live · Full results: firmulate.com/benchmarks.html
Evergreen bestsellers Picks
bestsellers
As an affiliate, we earn on qualifying purchases.
