It seems that Claude has been mass banning accounts. A friend of mine has already been banned three times. Every time it’s the same process: create a new Gmail, use an overseas phone number for verification, change to a clean IP, spend half a day registering successfully, use it cautiously for a few days, and then get banned again. Now he’s starting his fourth registration. But looking at social media, not a single person has said, "Forget it, let’s just use ChatGPT." You kick the users out, and they still want to come back. This might be the most magical product moat of 2026. Of course, Anthropic might not have the time to manage this. They’ve had a very exciting month. Looking back, at the end of February, the Pentagon wanted AI companies to sign contracts for any legal use, including autonomous weapons and mass surveillance. Anthropic refused and was labeled with "supply chain risk," a label previously only used for Chinese and Russian companies. OpenAI signed just a few hours later. At that time, the uninstallation rate of ChatGPT surged by 295%, and 1.5 million people joined the QuitGPT movement, while Claude shot to the top of the App Store. These events might be known to everyone. But looking back a month later, what’s truly interesting isn’t the drama at the time, but the results. Google won. Yes, not OpenAI, not Anthropic. It’s Google, which was hardly mentioned throughout the entire drama. During this time, Gemini had no resistance movement, no courtroom drama, no employee petitions. Quietly, they signed another contract with the Pentagon to provide AI workflows for 3 million government employees. I just saw the latest data: Gemini’s paid users increased by 258% year-on-year, surpassing Claude’s 200%. Google, with the bundling of Gmail, Docs, and Sheets, is quietly capturing the market on a completely different dimension. A month has passed, and all three have "won," but what they won is completely different. OpenAI won money. $25 billion in annualized revenue, $110 billion in financing. But they won’t be profitable until 2030, and one of the hottest posts on Reddit is still teaching people how to migrate chat records to Claude. Anthropic won reputation. Their enterprise API share surpassed OpenAI to 60:40, with $19 billion in annualized revenue, expected to be profitable two years earlier than OpenAI in 2028. But the CFO said in court that they might lose billions in 2026. Moreover, they are fighting legal battles in the U.S. to gain trust while simultaneously banning users in China. Google won position. The best AI in 2026 might not be the one with the highest score, but the one already embedded in the tools you can’t live without....