Marc Benioff, the Salesforce, Inc. CEO who has used ChatGPT every day for three years, said he would effectively stop using ChatGPT after trying Google's new artificial intelligence (AI) model "Gemini 3." As the global AI race is being reshaped every month, the public "pivot" by a leading Silicon Valley executive is drawing industry attention.
On the 24th (local time), Benioff wrote on X (formerly Twitter), "I used ChatGPT every day for three years. I tried Gemini 3 for two hours, and I'm not going back." He said, "Everything from reasoning and speed to image and video processing has become faster and sharper," adding, "It feels like the world has changed again." As of the morning of the 26th, the post had reached more than 3.2 million people.
His comments are drawing attention not just for a personal use experience but for their potential ripple effects across the industry. Forbes estimates Benioff's asset at about $8.5 billion, and Salesforce, Inc. has built close partnerships with major AI companies. Last month, it announced an expansion of its strategic collaboration with OpenAI and said it would integrate ChatGPT and the GPT-5 model into its platform "Agentforce 360." At the time, he emphasized, "corporations can offer the same immediacy and intelligence as ChatGPT to consumers."
The fact that Benioff moved to back Google's model over ChatGPT immediately after the Gemini 3 reveal is seen as showing how quickly the pecking order in AI technology can flip.
Google and DeepMind released Gemini 3 last week, calling it "the most intelligent model." Right after its debut, the model topped the LMArena leaderboard, an AI performance evaluation platform. Silicon Valley figures also offered positive reactions in succession. OpenAI CEO Sam Altman called it "a great model," and former Tesla AI director Andrej Karpathy said it is "a very solid tier-1 model for everyday use." Stripe CEO Patrick Collison also shared an interactive page summarizing "genetics innovations" generated by Gemini 3, calling it "astonishing."
Inside OpenAI, however, there are signs of alarm. According to The Information, Altman told employees in an internal memo that "a tough period may be coming" and that "Google has recently been posting impressive results across the board." He acknowledged that Google's surge "could create economic headwinds for the company in the short term," while stressing that "OpenAI is catching up quickly."
In recent days, the AI market has grown even more competitive as major models have launched in quick succession. Less than a week after Google unveiled Gemini 3, OpenAI released GPT-5.1, and Anthropic announced "Claude Opus 4.5" on the 25th. The race for leadership in AI capabilities is once again turning into a speed battle.