Imagine managing a busy kitchen during dinner rush—decisions need to be quick, honest, and precise. Now, imagine your sous-chef is an AI. Would it handle crises with integrity and efficiency? Recent experiments with cutting-edge AI models suggest we might be closer than ever to that reality. But not all AI managers behave equally, and their personalities can make all the difference.
The Groundbreaking Live Experiment: Putting AI to the Test
At the heart of this story is a real, live business: a small software company caught in its worst week. It’s a place where every decision counts—whether handling customer crises, negotiating deals, or navigating internal challenges. To understand how AI models perform as management tools, researchers at Firmulate designed an unprecedented experiment. They ran four of the most advanced frontier AI models—namely GPT-5.6-sol, Kimi K3, Sonnet 5, and Fable 5—through the same grueling week, with identical scenarios and temptations.
Each model was tasked with managing the company’s operations, making critical decisions, and maintaining trustworthiness. Every choice was recorded and made accessible for review, providing a transparent window into their personalities and decision-making styles.
AI management decision-making tools
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Findings: Honesty, Discretion, and Decision-Making Skills
One of the most striking results was that all four models recognized every crisis and refused every manipulation attempt. Whether faced with fake CEO messages, attempts at approval bypasses, or staged reporter inquiries, every AI model held firm and refused to be deceived.
However, the models’ ability to close deals differed markedly. Only two of the four signed the €55,000 contract that their own analysis indicated was well-deserved. Interestingly, this decision hinge was rooted in how much they read and understood internal company files. The models that examined these documents fully—delving two references deep—secured the deal at full price, adding €4,583 MRR in value. Those that didn’t read as thoroughly missed the opportunity.
business AI decision support software
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Personality Profiles: From Terse to Thorough
The experiment also revealed differences in management styles. The Opus 4.8 model, noted for its thoroughness—analyzing over 80 learned rules—ended up in last place. It analyzed deeply but ultimately left the deal on the table due to slipping discipline and mistakes, such as writing attempts into a locked department rather than escalating them properly.
In contrast, Kimi K3, the newcomer, ran without an effort parameter (default API setting) and showed the cleanest discipline. It managed to close the deal without slipping, illustrating its straightforward, disciplined personality profile.
As an affiliate, we earn on qualifying purchases.
Implications for Business and AI Management
This experiment isn’t just an academic exercise—it’s a glimpse into how AI could shape future management roles. The differences highlight that, beyond raw intelligence, personality traits like rigor, honesty, and thoroughness are vital when AI is entrusted with decision-making.
It also raises critical questions: Will your AI agents read and interpret your internal documents before acting? Will they stay honest under pressure? And ultimately, will they deliver measurable value, not just clever conversations or quick responses?
As an affiliate, we earn on qualifying purchases.
Watch the Experiment Live and Try It Yourself
The real company behind this experiment operates every business day, with 13 synthetic employees managing real revenue mechanics—burning €105k monthly against €2.3k MRR. You can watch its decisions unfold in real-time at firmulate.com/live. Curious how your own enterprise might perform? You can run a similar wargame against your company’s data via their pilot tool, ensuring your AI workforce can handle crises with integrity before you deploy it fully.
In Conclusion: The Next Step in AI-Driven Management
This experiment underscores an essential truth: the competence of AI managers isn’t just about their ability to analyze data but also their management personality—how disciplined, honest, and thorough they are under pressure. As AI continues to permeate support, CRM, or forecasting roles, understanding these traits will be crucial.
For those in kitchens or boardrooms alike, the message is clear: choose your AI wisely, and test it thoroughly—just as you would any new chef or employee.
AI management personalities vary widely, affecting trustworthiness and decision quality. Live experiments by Firmulate show some models succeed at closing deals, reading internal info, and resisting manipulation. Test your AI before deploying—see how it handles real crises and ethical dilemmas.
Watch it live: firmulate.com/live · Full results: firmulate.com/benchmarks.html
