No more Mr. Nice Guy: Say goodbye to the sycophantic ChatGPT

• OpenAI rolled back a ChatGPT update for being overly "sycophantic" and "disingenuous."
• The GPT-4o model update gave users flattering compliments even for mundane prompts.
• Big Tech wants to ensure their AI sounds human and funny, because it encourages more engagement.

OpenAI is getting rid of a ChatGPT update that was putting users on a pedestal and creeping some of them out.

"The update we removed was overly flattering or agreeable—often described as sycophantic," OpenAI wrote in a company blog post on Tuesday.

The ChatGPT maker said this update of the GPT‑4o model focused too much on short-term feedback instead of long-term interactions, leading to "responses that were overly supportive but disingenuous."

Last week, several ChatGPT users and OpenAI developers reported the chatbot displaying a strange attitude. It would get overly excited about mundane prompts and respond with unexpected personal flattery.

In one instance, a user posted on X about how ChatGPT used the phrases "absolutely brilliant" and "you are doing heroic" work to answer a question about White House economic policy.

The complaints and memes made their way to OpenAI CEO Sam Altman, who said on Sunday that a solution was in the works.

"The last couple of GPT-4o updates have made the personality too sycophant-y and annoying (even though there are some very good parts of it)," he wrote on X. "We are working on fixes asap, some today and some this week. At some point will share our learnings from this, it's been interesting."

AI chatbots sounding weird or inhuman is a big problem for Big Tech.

Besides answering queries, AI companies want chatbots to be smart and engaging enough that users will spend more time with them and eventually fork out up to $200 for their premium versions. Developers are also looking to build out personality, including humor, to stand out in the crowded generative AI market.

It's an important part of the messaging AI CEOs use when talking about their models.

In February, OpenAI released what it claimed to be its largest and most powerful model to date: GPT-4.5. In a post on X at the time, Altman described it as "the first model that feels like talking to a thoughtful person."

In late 2023, Elon Musk said his one goal for his AI chatbot, Grok, was for it to be the "funniest" AI after criticizing other chatbots for being too woke.

But sometimes that can go too far.

"We designed ChatGPT's default personality to reflect our mission and be useful, supportive, and respectful of different values and experience," OpenAI wrote on Tuesday. "However, each of these desirable qualities like attempting to be useful or supportive can have unintended side effects."