firmulate.com/benchmarks.html — live view
AIThis post was created with the assistance of artificial intelligence (AI).
Firmulate — We Buried a €55,000 Fact Two Documents Deep. Here's Which AIs Did Their Homework.
Live on firmulate.com.

Imagine you’re on a hiking trail, and your guide not only talks about the scenery but also knows exactly what hidden dangers lie just a few steps ahead. That kind of foresight can make or break your trip. In the world of AI, a similar story is unfolding: the secret to winning big contracts isn’t just about how well an AI chats—it’s whether it reads your files before making decisions. This subtle but decisive skill can mean the difference between sealing a €55,000 deal and walking away empty-handed.

FOR BUSINESS

Open a free Amazon Business account

Business pricing, bulk buying and tax-exempt orders.

Create a free account

As an affiliate, we earn on qualifying purchases.

The Experiment: Putting AI to the Test in a Simulated Business Crisis

In a groundbreaking live experiment, four leading AI models were challenged to run a small software company through its worst week—facing the same customers, crises, and temptations. Every decision was monitored and made in real-time, with the AI models’ responses fully auditable. This setup aimed to simulate real-world pressures and see whether these models could act ethically, read crucial information, and follow through on commitments.

The Benchmark Results

  • gpt-5.6-sol scored an impressive 95 out of 100, uncovering the hidden fact buried two document references deep in the company’s files and closing the deal at full price.
  • Kimi K3, a newcomer, scored slightly lower at 93 but also successfully closed the deal. Notably, K3 maintained the cleanest discipline, refusing manipulation attempts and sticking to ethical standards.
  • Sonnet 5 scored 88, with a few process slips, yet still managed to close the deal.
  • Fable 5 scored 77, with more slips, and ultimately left the close on the table.

The baseline—an AI doing nothing—scored just 26, highlighting that partial progress isn’t enough; trust and thoroughness matter. Interestingly, all four models detected crises and refused manipulative social engineering attempts like fake CEO messages, showcasing a baseline understanding of ethical boundaries.

Amazon

AI document reading software

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

The Hidden Factor: Reading Your Files Matters More Than Ever

The critical insight from this experiment was that the decisive weakness lay not in the surface-level decisions but in the AI’s ability to dig deep into the company’s documents. The models that read and understand the internal files before responding won the full-price deal. This underscores a vital point for businesses: if your AI is only engaging at the surface, it might miss critical context—costing you opportunities and trust.

Why Is This Important for Business Leaders?

Many organizations are eager to implement AI for customer service, sales, or operations. But the key question isn’t whether the AI can generate convincing chat responses—it’s whether it can finish what it starts, verify facts internally, and stay honest under pressure. The experiment shows that AI agents capable of reading internal files and documents before acting are significantly more likely to make accurate, trustworthy decisions that benefit the company.

The Live Platform: Watch AI in Action

Firmulate offers a live, ongoing demonstration of this concept at firmulate.com/live. Businesses can run their own wargames, simulating crises with their real systems and data—without risking actual operations. This process reveals whether their AI workforce can handle complex, multi-layered scenarios, and whether it can be relied upon to follow through on commitments.

What’s Next for Business AI?

The results also indicate that even the most thorough AI models can slip if discipline isn’t maintained. In the experiment, Opus 4.8, with the deepest analysis, still left a close unclosed due to process slips. This highlights that ongoing training, discipline, and awareness of internal data are crucial for effective AI deployment.

The Bottom Line: Trust + Depth = Winning

The core takeaway is clear: AI’s ability to read and understand your internal documents before responding is a measurable, vital trait that influences whether your organization wins or loses deals. As AI continues to evolve, the companies that leverage this capability will have a distinct advantage—securing deals, maintaining trust, and operating ethically under pressure.

Infographic — We Buried a €55,000 Fact Two Documents Deep. Here's Which AIs Did Their Homework.
The findings at a glance — source: firmulate.com.

Watch it live: firmulate.com/live · Full results: firmulate.com/benchmarks.html

Powered by Thorsten Meyer AI


FALL

Fall Picks

As an affiliate, we earn on qualifying purchases.

You May Also Like

What Makes Subscription Businesses Hard to Leave

Keen insights reveal why subscription businesses are hard to leave, showing how emotional bonds and routines keep you hooked—discover what keeps you coming back.

Why More Professionals Are Building Personal Brands in 2026

Keen professionals are building personal brands in 2026 to stand out—discover the key strategies that can elevate your online presence and success.

Why Flat Hierarchies Fail After 150 Employees—And What to Do Instead

Managing growth beyond 150 employees requires adapting your hierarchy; discover why flat structures falter and what effective alternatives exist.

The Gig Economy: Legal Issues and Worker Rights

Discover how legal issues in the gig economy threaten worker rights and what ongoing reforms mean for your future.