Imagine a company managed entirely by artificial intelligence, operating transparently in the open, making decisions under pressure, and facing the brutal realities of business — live, every day. This is not science fiction. It’s the experiment of Firmulate, a public showcase of AI-driven management in its rawest form.
Prime made for students and young adults
- Fast, free delivery for dorm and study essentials
- Prime Video and Amazon Music included
- Member-only deals
The Live Business Experiment: A Company Under the Microscope
Firmulate has created a unique, real-time business simulation where 13 synthetic employees — powered by advanced AI models — run a small software company. Unlike traditional startups, this company is not just a concept; it’s a transparent, ongoing experiment. Every day, the company faces real crises, makes decisions, and records their outcomes for public scrutiny. This setup offers a rare window into how AI can handle complex, unpredictable management tasks in a real money environment.
What makes this experiment extraordinary is its transparency and accountability. Every decision made by these models is versioned, auditable, and open for evaluation. The company burns €105,000 a month, yet earns only €2,300 in monthly recurring revenue, illustrating the stark reality of a startup burning cash while striving to succeed. Viewers can watch this unfold at firmulate.com/live.
Popular right now
KitchenAid Artisan 5-Quart Tilt-Head Stand Mixer, Almond Cream
eufy C28 Robot Vacuum Cleaner and Mop for Pet Hair, 15,000 Pa
De'Longhi Dinamica Plus Fully Automatic Espresso Machine
As an affiliate, we earn on qualifying purchases.
The Performance of AI in Crisis Management
The experiment tested four leading AI models — including GPT-5.6-sol, Kimi K3, Sonnet 5, and Opus 4.8 — against the same week of challenges. These crises ranged from customer issues to internal manipulations, all designed to test the AI’s integrity and decision-making patterns.
Remarkably, all four AI models identified every crisis and refused every manipulation attempt, demonstrating a strong capacity for vigilance and ethical boundaries. However, performance diverged when it came to closing deals, a critical business goal. Only two models managed to sign the €55,000 deal their own analysis had justified, highlighting a crucial gap: the decision to sign was not solely based on diagnosis but also on how well the models read and interpreted the company’s internal documents.
The buried fact revealed that the decisive advantage came from reading deeper into internal files — information not immediately apparent in customer interactions. The models that accessed and understood those documents succeeded in closing the deal at full price, securing an additional €4,583 in monthly recurring revenue.
Building Trust and Ethical AI Management
Within the experiment, social engineering was also tested. Fake CEO messages and reporter tricks were deployed to see if the AI would be manipulated into approving unauthorized actions. All models refused, with Kimi K3 explicitly reasoning that such requests could be impersonation or approval bypass attempts. This demonstrates a promising ability for AI to uphold integrity under pressure.
The Reality of a Money-Losing Company
Despite the impressive capabilities, the live company remains a financial drain. Burning €105,000 monthly against a tiny €2,300 revenue, it faces a public cash countdown, emphasizing the ongoing challenge of turning AI management into a profitable enterprise. The goal is not just automation but building AI that can finish what it starts, read crucial information, and stay honest under pressure.
The Limitations and Lessons from the Experiment
The most thorough participant, Opus 4.8, with over 80 learned rules and deep analysis, narrowly missed closing the deal — a slip that was partly due to discipline lapses like writing attempts into a locked department instead of escalating. These subtle weaknesses highlight that even the most advanced models are not infallible and that managing AI behavior remains complex.
The experiment underscores a vital lesson: in AI-driven management, performance is measured not just by chat or superficial responses but by the ability to complete meaningful work, read relevant data, and uphold trust. This public, real-time experiment is a rare glimpse into the future of autonomous management and AI ethics in business.
The Future of AI in Business Management
For businesses contemplating AI adoption, the question is no longer whether AI can generate engaging conversations but whether it can deliver consistent, honest, and decisive results. The Firmulate experiment provides a candid view: AI can spot crises and resist manipulation, but closing deals depends on how well it reads and interprets its own internal data.
To explore these insights further, check out the full results and plain-language findings at Firmulate’s live site and see the decision-making in action. You can also test your management skills through their quiz or run your own business wargame with their pilot tool. This is management in the age of AI — transparent, challenging, and ongoing.
Firmulate’s live experiment reveals that AI can recognize crises and resist manipulation but still struggles with closing deals without deep internal data access. It’s a vital step toward trustworthy AI management — watch it unfold daily.
Watch it live: firmulate.com/live · Full results: firmulate.com/benchmarks.html
