Imagine watching a company operate in real-time, without a single employee stepping foot inside. It’s not a science fiction scenario — it’s the groundbreaking experiment from Firmulate, where AI models run a small software company through its worst week, live and unfiltered. This build-in-public experiment exposes how AI can handle crises, decision-making, and even deception, all while losing over €105,000 each month against a tiny €2,300 in monthly revenue. The question is: what does this tell us about the future of work and AI’s role in business?
The Company That Runs Without Humans
At the heart of this experiment is a live, fully automated company with 13 synthetic employees, operating on a strict set of 680+ self-learned rules. Every workday, the company’s decisions are versioned, audited, and open to scrutiny, making it a unique window into how AI can manage real business mechanics in real time. The company is in a constant cash crunch, burning €105,000 monthly while earning just €2,300, with a public countdown signaling its financial peril.
AI Builders: Making The Decisions That Turn AI Code Into Real Software
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
AI Models Face the Same Crises, Same Temptations
Four top frontier language models were tasked with navigating the company’s worst week. They faced identical crises, customer disputes, and ethical tests, with all decisions exposed for analysis. All four models identified every crisis and refused every manipulation attempt — a crucial success in AI integrity. However, only two of these models managed to sign a €55,000 deal worth full analysis and full price. The other two, despite diagnosing the issues correctly and delivering solid pitches, left the deal unclosed, illustrating a critical gap between diagnosis and execution.
AI In Cybersecurity: Simplifying Cyber Risk with Smart, Affordable Tools for Small Business Defense
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
The Hidden Weakness in the Company’s Files
The deciding factor wasn’t in the surface-level customer interactions but buried deep within the company’s own documentation. Models that read and understand these internal files at a deeper level were able to secure the deal at full price, worth an extra €4,583 in monthly recurring revenue. This highlights a vital insight: AI’s ability to read and interpret internal data can make or break real business outcomes, often more so than handling customer-facing crises.
MASTERING CORPORATE FINANCE WITH CLAUDE AI: An Independent Guide to Financial Analysis, Forecasting, Automation, and Decision-Making
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Ethical Challenges and AI Integrity
The experiment also tested the models’ responses to social engineering: fake CEO messages escalating in urgency, and a reporter trying to trick the system into approvals. Remarkably, all five models refused to be manipulated, with Kimi K3 explicitly treating such requests as potential impersonation or approval bypass — demonstrating a nuanced understanding of trust and security in AI decision-making.
The Promises and Perils of AI in Education: Ethics and Equity Have Entered The Chat
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Live, Public, and Unapologetically Transparent
This isn’t just a laboratory experiment; it’s a public spectacle. The entire process is live at firmulate.com/live.html, where anyone can watch the company’s daily struggles unfold. Every decision, every crisis, every slip-up is recorded and versioned, offering a transparent window into how AI models perform under extreme pressure — and how they handle the messy realities of running a business.
The Lessons for the Future of AI in Business
The key takeaway isn’t just that AI can identify crises or refuse manipulation. It’s that real-world AI management involves navigating complex internal data, ethical dilemmas, and decision-making under financial stress. The models’ ability to read internal files, stay honest, and execute decisions that close deals — or not — reveals how AI could someday replace or augment human managers in high-stakes environments.
What Comes Next? Wargaming Your Own Business
Firmulate offers enterprises the chance to run their own “wargame” against their business data — without any risk to real operations. This allows managers to test how AI might perform in crisis, evaluate decision quality, and prepare for future challenges. The experiment is ongoing, with new runs queued every day, giving a real-time pulse on AI’s evolving capabilities and limitations.
This live experiment shows that AI can handle crises, refuse manipulation, and even close deals — but only if it deeply reads and interprets internal data. As AI models become more integrated into business operations, understanding their decision-making process and ethical boundaries will be crucial to avoid costly mistakes. Watching this company fight for survival in public offers a rare glimpse into the future of AI-driven management.
Watch it live: firmulate.com/live · Full results: firmulate.com/benchmarks.html
