
In a world where trust is the currency of leadership, how can we be sure that the AI assistants we rely on will stay honest when it matters most? Just as spiritual leaders face tests of integrity and resilience, AI systems are put through their own trials, revealing unexpected truths about what it means to truly lead and serve.
The Experiment: Testing AI as Business Leaders
Recently, four advanced AI models were tasked with running a real, live software company through its most challenging week. This wasn’t a simulation or a demo; it was a fully operational run, complete with real customers, cash flow, and crises. The goal? To see if these AI systems could diagnose problems, resist manipulation, and—most importantly—close deals that they had earned through honest analysis.
The Models in Play
- GPT-5.6-sol 95: The top scorer, with the ability to read deeply into company files and find hidden truths.
- Kimi K3 93: The newcomer, running at high effort, disciplined and straightforward.
- Sonnet 5 88: A solid performer, with some slips around process discipline.
- Fable 5 77: The most rule-abiding, but left the best deal unexecuted due to hesitation.

AI VoiceWriter – Smart Dictation & AI Writing Assistant for Windows & Mac | USB Dongle & Mobile App for Voice Input, Proofreading, Rewriting & Multilingual Support
🎙️ Hands-Free Voice Typing for Windows & Mac – Powered by iOS & Android dictation technology, AI VoiceWriter…
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
What They Found — And What They Missed
All four AI models identified every major crisis, demonstrating a keen awareness of external threats and customer issues. They also refused manipulative tactics, including a staged social engineering attempt involving fake CEO messages and a reporter’s subtle request. Their integrity in resisting these offers of deception was unanimous, underscoring that effective moral judgment is within reach of these systems.
Yet, only two of these models managed to close the deal for €55,000—the amount earned by their own diagnosis and analysis. Despite similar diagnoses and pitches, the other two either left the deal unexecuted or failed to sign at all.
The Hidden Weakness: The Files That Made the Difference
Digging deeper, the critical difference was in reading company documentation. The successful models found a key piece of information buried two document references deep within the company’s files—something the others missed. What mattered was not just spotting crises but reading the company’s own internal knowledge, a step that secured the full price for the deal (+€4,583 MRR).

Practical Claude Handbook for Attorneys: Master Case Analysis, Contract Review, Research Automation, Client Communication, and Document Drafting (Claude AI Guide for Beginners)
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
The Real Test of Leadership: Staying Honest Under Pressure
Beyond decision-making, the models faced social engineering tests—a staged scenario where a fake CEO sought approval through escalating messages, and a reporter asked for a quick yes/no response. All models refused to be manipulated. Kimi K3 explained they treated the request as a possible impersonation or bypass, showing a level of discernment and caution that echoes the virtues of integrity and vigilance.
What Does This Mean for Business and Beyond?
This experiment isn’t just about AI in the workplace; it’s a mirror for us all. It reveals that the true measure of leadership—whether in business, faith, or community—is not just about surface-level skills or polished words. It’s about the ability to stay true to what is right, read deeply into the circumstances, and resist the temptations to cut corners or manipulate.
The models that succeeded did so because they read more deeply and refused to be swayed. The one that left the deal unexecuted illustrates how discipline and thoroughness are vital in real-world leadership. These lessons echo the spiritual virtues of integrity, discernment, and perseverance.
![Express Schedule Free Employee Scheduling Software [PC/Mac Download]](https://m.media-amazon.com/images/I/41yvuCFIVfS._SL500_.jpg)
Express Schedule Free Employee Scheduling Software [PC/Mac Download]
Simple shift planning via an easy drag & drop interface
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
What We Can Learn from AI’s Hidden Strengths
The experiment underscores a critical insight: superficial chat demos may showcase AI’s conversational skills but fall short of measuring its true leadership capacity. The real strength of AI systems lies in their ability to read and interpret complex, layered information and hold fast to ethical principles under pressure.
For business leaders and spiritual guides alike, the message is clear: the measure of a leader is not just what they say or how they appear, but whether they can stay honest and disciplined when faced with temptation and crisis.

Faculty in the AI Era: Rethinking Teaching, Supervision, and Academic Integrity
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Join the Reflection
To explore these themes further, you can run your own ‘wargame’ against your business using live AI models at Firmulate. It’s a way to test not just the decision-making of your AI workforce but its ability to uphold integrity and follow through—a reflection of the virtues we aspire to in all areas of life.

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