OpenAI declares 'code red' as Google's Gemini, Anthropic gain ground on ChatGPT

OpenAI’s boss told staff Monday the company is hitting the panic button on ChatGPT. Sam Altman sent a memo declaring “code red” and said other projects are getting pushed back so the team can fix the chatbot.

The Wall Street Journal saw the internal message. Altman laid out what needs work: better personalization, faster speeds, and more reliable performance. He also wants ChatGPT to handle more types of questions.

This is the biggest signal yet that OpenAI feels the heat from competitors. Google worries Altman the most. Last month, Google rolled out a new Gemini AI model that beat OpenAI on industry tests, as reported by Cryptopolitan. Google’s stock jumped after that.

Gemini’s been picking up users fast since August, when it launched Nano Banana, an image generator. Google says monthly active users went from 450 million in July to 650 million by October. Then there’s Anthropic, which is winning over business customers.

Billions in data center spending create pressure

OpenAI has committed to spend hundreds of billions on data centers down the road. Investors are getting nervous about when that money will actually pay off. The company’s still private, CFO Sarah Friar said back in November at a Journal event that an IPO isn’t happening anytime soon. But what happens to OpenAI matters a lot to Nvidia, Microsoft and Oracle.

Altman’s memo said work on other stuff is getting delayed. That includes advertising, AI tools for health and shopping, and something called Pulse that was supposed to be a personal assistant.

He’s encouraging people to switch teams temporarily and said there will be daily calls for everyone working on ChatGPT fixes. Monday night, Nick Turley who runs ChatGPT posted on X that the focus now is growing the chatbot and making it “even more intuitive and personal.”

No profit path without massive growth

Here’s the thing about OpenAI: it’s not making money. The company has to keep raising funds just to stay alive. That puts it at a disadvantage against Google and other big tech firms that can pay for things with revenue they already have. OpenAI spends more aggressively than Anthropic too. Based on OpenAI’s own numbers, it needs to hit roughly $200 billion in revenue to turn a profit in 2030.

Altman’s kept financial worries at bay mostly because ChatGPT has a massive user base. More than 800 million people use it weekly. Plus OpenAI’s still ahead in cutting-edge AI research. In his memo, Altman said a new reasoning model coming out next week will beat Google’s latest Gemini. He claims the company’s doing well on other fronts too.

OpenAI’s had a tough time finding the right mix between keeping the chatbot safe and making it interesting to use. When GPT-5 dropped in August, lots of users weren’t impressed. The complaints? It came across too robotic and struggled with basic stuff like simple math problems and geography facts.

OpenAI rolled out an update in November to fix the tone and make it better at doing what users actually asked for.

Turns out the company had already sounded the alarm before Monday’s announcement. They’d called a “code orange” over ChatGPT’s problems, the memo showed. OpenAI runs on a system with three alert levels, yellow sits at the bottom, orange in the middle, and red at the top, based on how serious the issue is, according to people who know how it works.

If you're reading this, you’re already ahead. Stay there with our newsletter.