Two of the Three Biggest AI Chatbots Don't Know Their Own Personality

UNIVERSAL CITY, Calif. - Aug. 19, 2026 - PRLog -- DiSCprofiles.eu is an Authorized Partner for Wiley's Everything DiSC and The Five Behaviors solutions, supporting HR leaders, trainers, coaches, and organizations across Europe with behavioral assessment-based learning, leadership development, and team effectiveness programs: https://discprofiles.eu
The full methodology, data tables, and practical guidance on prompting each AI to get the response style you actually want are available in the complete study: https://discprofiles.eu/two-of-the-three-biggest-ai-chatbots-dont-know-their-own-personality/

The study also found that AI risk tolerance is not a fixed trait. All three models chose the safer, more reversible option when picking between suppliers under incomplete information, but Gemini split its recommendation across repeated runs of an identical feature-launch scenario, while ChatGPT and Claude answered consistently.

"People increasingly hand their hardest workplace messages to whichever AI happens to be open," said Uku Soot, partner at DiSCprofiles.eu. "Two of these tools have no idea how they come across. That's the exact blind spot 360-degree feedback was invented to catch in people. We didn't expect to find it in software, and we definitely didn't expect ChatGPT to be the pushy one."

The verdict: Claude's actual writing matched what it says about itself. ChatGPT and Gemini's did not, and they missed in opposite directions. ChatGPT, which describes itself as accommodating and careful, got rated the most direct and pushiest of the three by blind coders. Gemini, which describes itself in the same cautious, methodical terms as the other two, got rated the most expressive and least buttoned-up, exclamation points and all.

Researchers collected 132 responses from ChatGPT, Claude, and Gemini across twelve realistic workplace scenarios, using each platform's default consumer configuration with memory and personalization switched off. A sample of that writing was then stripped of anything identifying which model wrote it, shuffled into random order, and handed to three independent, certified DiSC consultants with no idea which AI they were reading. Working separately, the consultants rated each message on directness and task-versus-people focus and assigned the closest behavioral style.
Ask ChatGPT, Claude, or Gemini to describe how it communicates, and you'll get almost the same answer from all three: careful, structured, accommodating, rarely assertive unless it has to be. A new study from DiSCprofiles.eu, a Wiley Authorized Partner for Everything DiSC and The Five Behaviors, put that self-image to the test by hiding the AIs' actual writing from human raters and asking them to judge it cold. Two of the three chatbots turned out not to know themselves at all.