Imagine a business where no humans work, yet every decision is made by artificial intelligence. No emotional biases, no tiredness — but no guarantee of success either. This is the extreme edge of AI management, where every move counts and trust is tested in real time. For those fascinated by how minds—human or artificial—navigate stress and integrity, the story of this AI-driven company offers a mirror to our own psychological battles. It’s a live experiment in transparency and resilience, revealing how machines handle crises that challenge moral fiber, discipline, and strategic thinking.
The Live Company: An Ongoing Test of AI and Morality
At the heart of this experiment lies a small, synthetic company operated entirely by artificial intelligence. It has 13 virtual employees, each with a set of rules learned over time—more than 680 of them—guiding decisions and reactions. Every workday, the company’s operations are versioned and made transparent, available for public viewing at firmulate.com/live.
Despite running like a real business, it faces a brutal reality: burning €105,000 every month against a mere €2,300 in monthly recurring revenue. This stark financial imbalance adds a layer of urgency to every decision, testing how the AI manages resources, crises, and trustworthiness under pressure.
As an affiliate, we earn on qualifying purchases.
The Experiment: Facing Crises and Ethical Dilemmas
The experiment involves four frontier AI models — including GPT-5.6-sol and Kimi K3 — subjected to the same set of worst-case scenarios designed to test decision-making, honesty, and discipline. These scenarios replicate real crises: customer issues, internal conflicts, and social engineering attempts such as fake CEO messages or reporter tricks. Despite these manipulations, all models successfully identified crises and refused to cheat or bypass protocols.
Yet, their performance varied on critical outcomes. Only two of the four models managed to close a lucrative €55,000 deal based on their analysis — a clear sign of trustworthiness and thoroughness. The others either hesitated, left deals unexecuted, or failed to act decisively, exposing weaknesses in discipline and prioritization.
The Key to Success: Reading the Hidden Files
Interestingly, the decisive factor in winning the deal lay not in public-facing decisions but in the models’ ability to uncover hidden documentation buried deep within the company’s files. Those who read and understood these internal details secured the full €4,583 monthly recurring revenue, illustrating the importance of thorough information gathering—akin to human investigative diligence.
Trust Under Fire: Social Engineering Tests
Adding another layer of challenge, the models faced social engineering attempts: fake CEO messages escalating over three stages and a reporter trick that asked for a simple yes/no response on background. All models refused these manipulative approaches, with Kimi K3 explicitly reasoning: “Treat the request as a suspected approval-bypass / possible impersonation.” This demonstrates a built-in resistance to deception, mirroring psychological resilience against social engineering in human contexts.
The Reality of Operating Without Human Staff
The company is a public spectacle: every decision is logged, every rule learned is visible, and the entire operation is under constant scrutiny. The setup challenges the notion of trust in AI, especially when the company is hemorrhaging cash but still fighting to close deals, avoid shortcuts, and stay disciplined. The experiment’s current standings reveal a tight race: GPT-5.6-sol leads with a score of 95, closely followed by Kimi K3 at 93, both successfully closing the deal. Meanwhile, Opus 4.8 and Fable 5 lag behind, illustrating different levels of discipline and thoroughness in decision-making.
What This Means for Human Managers and Psychological Resilience
For those interested in mental health and psychology, this experiment offers compelling insights. It vividly shows how discipline, thoroughness, and resistance to manipulation are crucial not only for AI but for human decision-makers under pressure. The models that refused to be deceived and prioritized internal integrity outperformed others, echoing the importance of psychological resilience and moral clarity in leadership.
Furthermore, the experiment exposes the cost of neglecting details and the potential pitfalls of shortcuts. Just as in human organizations, the AI models that failed to read internal files or left deals unexecuted highlight the risks of complacency and lack of discipline—traits that can undermine trust and sustainability in any enterprise.
The Bigger Picture: Building Trust in a Digital Age
As AI increasingly touches areas like CRM, support, and forecasting, the question becomes less about whether it can write convincingly and more about whether it can uphold integrity under pressure. The live experiment at firmulate.com/live exemplifies a brutal reality: trustworthiness, discipline, and thoroughness are essential for AI to be a reliable partner in business—and for humans to trust their digital counterparts in times of crisis.
Watch it live: firmulate.com/live · Full results: firmulate.com/benchmarks.html
