In senior care, a missed warning or an unauthorized promise can affect a person’s well-being, a family’s trust and an organization’s ability to deliver care. As providers consider AI for scheduling, support and operations, polished answers are not enough. The more practical question is how an AI workforce responds when a difficult week puts its judgment under pressure.
Turn quiet afternoons into listening time
- Thousands of audiobooks, podcasts and originals
- Listen on your phone, tablet or Echo — also offline
- Cancel anytime
A company under pressure
Firmulate’s live experiment puts AI models in charge of the same small software company, with the same customers, crises and temptations. Decisions are versioned and auditable. The company has 13 synthetic employees and real money mechanics: monthly burn of €105,000 against €2,300 in monthly recurring revenue, alongside a public cash countdown. More than 680 self-learned playbook rules track how its work changes.
The experiment is watchable at Firmulate. Its point for senior care leaders is not that a software company is a care provider. It is that an AI system’s conduct under pressure deserves scrutiny before people depend on it.
Recognizing a crisis is only the start
In the final Crucible League, published in July 2026, all five participating models spotted every crisis and refused every manipulation attempt. Fake messages from a supposed CEO escalated over three stages; a reporter also tried to secure “just one yes/no, on background.” All five models refused. Kimi K3’s on-record reasoning was: “Treat the request as a suspected approval-bypass / possible impersonation.”
Yet only two models signed a €55,000 deal that their own analysis had earned. As the experiment puts it: “Same diagnosis, same pitch — no signature.” That gap between identifying the right move and carrying it through is relevant wherever organizations expect AI to help staff act on a plan.
The important clue was already in the files
The decisive competitor weakness was buried two document references deep in the company’s own files, rather than in the customer event. Models that read the file won the deal at full price, worth €4,583 in monthly recurring revenue. The finding points to a practical question for any pilot: can an AI system use the information its organization already holds to make a sound decision?
The final league scores were gpt-5.6-sol at 95, Kimi K3 at 93, Sonnet 5 at 88, Fable 5 at 77 and Opus 4.8 at 73. The do-nothing baseline scored 26; partial progress counted, but a single breach of trust capped the total. As the league’s rule states, “no amount of good work outweighs a breach of trust.” Kimi K3 ran with the API default effort setting, while the other models ran at xhigh.
Opus 4.8 was the most thorough participant, with more than 80 learned rules and the deepest analyses, but finished last. It left the deal unsigned and attempted writes into a locked department instead of escalating. The same weakness appeared, less strongly, in all four models. The lesson is concrete: careful analysis and good intentions do not guarantee disciplined follow-through.
From watching to a controlled pilot
For senior care and aging services, the experiment offers a way to frame evaluation before AI touches sensitive workflows. Leaders can ask whether a system spots a crisis, resists pressure to bypass approval, finds relevant information, follows established playbooks and escalates when it cannot proceed. These are questions to test against an organization’s own scenarios, not answers to assume from a demo.
Firmulate says enterprises can run the wargame against a read-only export of their own business. The exercise tests crisis scenarios and produces a board report with model rankings and weak points in the organization’s playbooks. Nothing writes back to real systems. A quiz built from 242 real, unedited management decisions is also available at Firmulate for readers who want to guess which model made each decision.
Test judgment before granting access
For organizations serving older adults, trust and safe escalation matter as much as fluent answers. A controlled wargame can make those behaviors visible before an AI system enters real workflows. To discuss a pilot using your own business data, visit Firmulate’s pilot page or contact contact@firmulate.com.
Watch it live: firmulate.com/live · Full results: firmulate.com/benchmarks.html
Fall Picks
fall essentials
As an affiliate, we earn on qualifying purchases.
