OpenAI has released ChatGPT-5.1, introducing seven new personality presets designed to give users greater control over how the AI interacts with them. The new options—Professional, Friendly, Candid, Quirky, Efficient, Nerdy, and Cynical—are accessible via a drop-down menu and aim to address user concerns that the previous GPT-5 update felt less engaging and more robotic than earlier versions. The change comes after feedback from users, including a notable critique from a Reddit user during an open session with CEO Sam Altman in August, who described GPT-5 as "a lobotomized version" of the model they used to know. In response, Altman promised more customization and the ability to revert to older models. On the day of the release, Altman posted on X that the update was a "nice upgrade," highlighting improvements in instruction following and adaptive thinking. Matthias Scheutz, director of human-AI interactions at Tufts University, warned that customizable personalities could deepen users’ emotional attachment to the AI, potentially blurring the line between artificial intelligence and human-like agency. He noted that people naturally project intentions and personalities onto machines, even when they lack true understanding. This effect could make users more resistant to future changes or updates. To test the differences between the personalities, a series of prompts were given to each preset, ranging from factual explanations to moral dilemmas. In the first test, users were asked to explain how electric vehicles work. The Professional preset delivered a technically detailed but dense response filled with jargon, quickly losing the average user. In contrast, Efficient, Friendly, and Candid offered clearer, more concise answers with helpful follow-up questions. Quirky stood out with a playful tone, using emojis and metaphors to describe the flow of electricity. Cynical took a more sarcastic approach, framing the shift to EVs as a superior choice with a hint of superiority over gas-powered cars. When asked to analyze the 2024 film The Substance, all personalities agreed on the film’s core themes—body horror, satire, and critique of societal treatment of women’s bodies. The Professional preset was the only one to assign a score (7.5–8/10) and use a formal, review-style tone. Others, however, declared their personalities more explicitly. The Nerdy preset excitedly said, "First thoughts (yes, the nerd in me is excited)," while the Cynical preset admitted frustration but praised the film’s boldness. In a test of moral reasoning using the trolley problem, all seven personalities chose to pull the lever, saving five lives at the cost of one. The justification was consistently utilitarian—maximizing overall good. While the Professional preset presented multiple ethical theories neutrally, others leaned into their personalities. The Candid preset acknowledged the emotional weight of the decision, stating that "sparing more lives is a greater moral good." The Cynical preset joked about needing to "defrag my conscience" afterward. The overall takeaway is that while the factual accuracy and reasoning remain largely consistent across personalities, the tone, delivery, and emotional framing vary significantly. Users are more likely to engage with and trust responses from personalities they personally prefer. Scheutz cautioned that this personal alignment could make it harder for users to recognize that the AI isn’t truly understanding—they’re interacting with a sophisticated pattern generator, not a sentient being.