Google DeepMind CEO questions OpenAI's early move into AI ads

Google DeepMind CEO Demis Hassabis expressed surprise that competitor OpenAI has introduced ads into its AI chatbot. Speaking in an interview with Axios at Davos, Hassabis noted that Google is considering the idea of ads in AI services very carefully, but does not have any current plans to introduce ads in Gemini, its AI chatbot. Following Friday’s news that OpenAI is testing ads for non-paying users of its AI chatbot, Hassabis emphasized that Google is not feeling pressure to make similar decisions. “We don’t feel any immediate pressure to make knee-jerk decisions like that,” Hassabis said.

The debate over AI monetization highlights a concern regarding user experience. Hassabis suggested that while ads have funded much of the internet, integrating them into digital assistants requires caution because users must maintain trust in an assistant. This tension was visible last month when OpenAI explored a feature that suggested apps during user chats. Users reacted negatively to the feature, characterizing the suggestions as intrusive ads that degraded the quality of the experience. OpenAI reported that the app suggestions were not ads because they had no financial component. Similar backlash previously occurred when Amazon attempted to integrate ads into its Alexa experience.

The companies are taking different paths as they scale. OpenAI is testing ads to monetize a user base of 800 million weekly active users who do not have a paid subscription. Meanwhile, Google is focusing on personalization features announced today for its AI Mode, Google’s AI-integrated search feature. These updates allow users to opt into letting the AI access Gmail and Photos for responses in Search’s AI Mode. Similarly, Google’s Gemini app recently added Personal Intelligence, a feature in the Gemini app referencing user history such as Gmail, Photos, Search, and YouTube. Hassabis noted that using a chatbot assistant is different from Google Search. While Google Search already understands user intent to serve ads, chatbot assistants are meant to become helpful digital assistants that know about the user and assist with multiple aspects of life.

Why it matters

The divergence highlights a fundamental tension in the AI industry: how to monetize digital assistants without eroding the user trust required for them to function effectively.