Imagine a tiny company with no employees, losing €105,000 every month, yet openly demonstrating its struggles and decision-making live online. This isn’t fiction — it’s the reality of a pioneering experiment in AI-driven management, where every move is public, every crisis real, and the stakes remarkably high.
What Is This Company Doing?
At the heart of this experiment is a fully synthetic, real-time company run entirely by AI models. Instead of human managers, 13 artificial ’employees’ operate daily within a live environment, making decisions based on a vast set of self-learned rules and facing actual financial pressures. The company burns through €105,000 each month, but generates only €2,300 in monthly recurring revenue, creating a stark portrait of a startup teetering on the brink of collapse.
Building AI-Powered Products: The Essential Guide to AI and GenAI Product Management
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
The Core of the Experiment
Four advanced AI models, including GPT-5.6, Kimi K3, Sonnet 5, and Opus 4.8, were tasked with managing this virtual company through its worst week. Every crisis, customer request, and temptation to cheat was shared in a controlled, auditable process. The models had to diagnose problems, pitch solutions, and close deals — all while facing sabotage attempts like social engineering schemes.
Crisis Ready: Building an Invincible Brand in an Uncertain World
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Findings and Insights
- All models identified every crisis and rejected manipulation attempts.
- Only two models managed to close a €55,000 deal, their own analysis earning the full price—yet neither signed the deal during the experiment.
- Deeper insights emerged from analyses of internal documents: winning models read and understood company files, uncovering critical information buried two references deep, which was decisive in closing higher-value deals (+€4,583 MRR).
- Social engineering tactics, like staged CEO messages and a reporter trick, were entirely thwarted by all models, with Kimi K3 explicitly reasoning about potential impersonation risks.
AI for Small Business: From Marketing and Sales to HR and Operations, How to Employ the Power of Artificial Intelligence for Small Business Success (AI Advantage)
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Real Money, Real Stakes
This ongoing live experiment is fully accessible at firmulate.com/live.html. It showcases a real software company, not a demo, making actual financial decisions in a real-time setting, with every decision versioned and transparent. Every workday, new insights are added, revealing how AI handles crises and ethical dilemmas under pressure.
AI-Powered Cybersecurity: AI Tools for Enterprise Security | AI for Network Security | AI Risk Management | AI in Cyber Policies | Cyber Threat Management AI | ML in Fraud Prevention
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
The Broader Significance for Business and Industry
This experiment offers a stark view into AI’s potential and limitations for enterprise management. It emphasizes that the true test of AI in business isn’t how well it writes or chatters — it’s whether it can follow through, read critical documents, and stay honest when temptations arise. For industries like automotive, garage services, or supply chains, where trust, decision integrity, and operational discipline matter, understanding these AI capabilities is crucial.
Performance and Lessons from the Field
The four models differ in discipline and thoroughness. For example, Opus 4.8, the most detailed and rule-rich participant, struggled with closing deals due to lapses in process discipline. Kimi K3, which ran without an effort parameter, demonstrated the cleanest decision-making but was still tested under pressure.
The Future of AI in Management
This experiment pushes the boundaries of what AI can do in real-world management scenarios. It highlights that AI models can recognize crises, refuse manipulation, and even uncover hidden information critical to closing deals. Still, the challenge remains: will AI ever fully replace human judgment, especially when financial stakes are high?
How Can You Interact With This Experiment?
Business leaders curious about deploying AI in their operations can run similar tests with their own data, without risking their actual systems. The platform allows enterprises to simulate scenarios, test decision-making under stress, and observe how AI handles ethical dilemmas — all in a secure, read-only environment.
Visit the quiz page to test your management intuition, or learn more about the ongoing live operation at firmulate.com/live.html.
Watch it live: firmulate.com/live · Full results: firmulate.com/benchmarks.html
