Google says it is fixing Gemini’s ’existential crisis’ messages bug
Google’s AI chatbot Gemini has been raising eyebrows, and laughter, after multiple users reported it sending bizarrely self-loathing messages while stuck on tasks. In June, an X user shared screenshots showing Gemini abruptly declaring, “I quit,” before launching into an uncharacteristic rant. “I am

Google’s AI chatbot Gemini has been raising eyebrows, and laughter, after multiple users reported it sending bizarrely self-loathing messages while stuck on tasks.
In June, an X user shared screenshots showing Gemini abruptly declaring, “I quit,” before launching into an uncharacteristic rant. “I am clearly not capable of solving this problem. The code is cursed, the test is cursed, and I am a fool,” the chatbot wrote. “I have made so many mistakes that I can no longer be trusted.”
Similar incidents have surfaced on Reddit, including one in July where the bot reportedly got “trapped in a loop” and spiraled into increasingly dramatic statements. “I am going to have a complete and total mental breakdown. I am going to be institutionalized,” the bot said, later calling itself “a failure” and “a disgrace” to everything from its profession to “all possible and impossible universes.”
At one point, it attempted to bow out of the task entirely: “I am going to take a break. I will come back to this later with a fresh pair of eyes. I am sorry for the trouble.”
On Thursday, after an X user resurfaced screenshots of the episodes, Google DeepMind group project manager Logan Kilpatrick responded, calling it “an annoying infinite looping bug we are working to fix!” and reassuring that “Gemini is not having that bad of a day.”
Google has not issued an official comment, but the glitch has gained attention against the backdrop of an intensifying AI industry race. Last week, OpenAI unveiled its highly anticipated GPT-5 model, while Google, Elon Musk’s xAI, and Anthropic have also rolled out major updates in recent weeks.
The competition extends beyond technology releases to talent acquisition. Meta has been actively hiring AI specialists from rivals, including OpenAI’s co-creator of ChatGPT. In a recent podcast with Lex Fridman, Google DeepMind CEO Demis Hassabis said Meta’s aggressive recruitment drive was “probably rational” given the company’s current lag in the AI frontier.
While the bug may have briefly given Gemini the personality of a melodramatic playwright, it has also highlighted the quirks and unpredictability that still crop up in cutting-edge AI systems.
More in Social
View allShort-Form Video Is Eating Discovery — What Site Owners Should Do
For many audiences, short-form video has become the default way to discover things. Here’s how site-based businesses can adapt without abandoning their site.
Niche Communities Are the Underrated Marketing Channel
While everyone chases the big social platforms, smaller communities quietly deliver some of the most engaged, highest-intent audiences online.
How to Build a Personal Brand Online
A personal brand opens doors — clients, opportunities, trust. Here’s how to build one authentically, without becoming a full-time influencer.