Google AI Mode's Personal Context Features Delayed

Seven months after Google first teased "personal context" features for its AI Mode at Google I/O, the highly anticipated capabilities, including opt-in Gmail connections, are still not ready for public release. Nick Fox, Google's Senior Vice President of Knowledge and Information, confirmed that these features remain in internal testing.

Speaking in an interview with the AI Inside podcast, Fox clarified that the delay stems from product and permissions challenges, rather than limitations in the AI model's capabilities. "It's still to come," he stated, indicating ongoing development.

What Google Promised at I/O

During Google I/O, the company announced that AI Mode would "soon" integrate a user's past searches to deliver more relevant responses. A key promise was the ability for users to opt-in and connect other Google applications, starting with Gmail, with robust controls to manage these connections. The core idea was to eliminate the need for users to repeatedly provide context in every query, allowing Google to leverage relevant details already stored in their accounts.

Regarding a public launch date, Fox remained vague, noting only that:

"Some of us are testing this internally and working through it, but you know, still to come in terms of the public roll out."

AI Mode's Continued Growth Without Personal Context

Despite the absence of these personalization features, AI Mode has seen significant adoption. Fox highlighted its rapid expansion, reporting that AI Mode has "grown to 75 million daily active users worldwide."

Interestingly, the lack of personal context has influenced how users interact with the AI. Fox observed that queries are now "two to three times as long," with users explicitly adding first-person context into their prompts. Instead of relying on AI Mode to infer intent, people are manually providing the necessary background. As Fox put it:

"People are trying to put the right context into the query."

This behavior underscores the very problem the personal context feature was designed to solve: reducing manual effort.

Geographic Patterns in Adoption

Adoption rates for AI Mode vary geographically. Markets where AI Mode was introduced earliest, such as the U.S., show the strongest traction, becoming a more established part of users' routines. Fox described the U.S. as the most "mature" market.

Additionally, strong adoption is seen in regions where the web is less developed in certain languages or areas, including India, Brazil, and Indonesia. In these markets, AI Mode's ability to synthesize information across languages and borders offers a significant advantage over traditional search results. Younger users, across all regions, are also embracing the experience more quickly.

Publisher Relationship Updates

The interview also touched upon Google's efforts to connect users back to publisher content within AI Mode. One such feature is "Preferred Sources," which allows users to prioritize specific publications in Google's Top Stories unit. This feature is currently available worldwide in English.

Fox also detailed ongoing improvements to links within AI experiences, including increasing the number of links displayed and adding more contextual information around them. He stated:

"We're actually improving the links within our AI experience, increasing the number of them…"

On the commercial front, Google has established partnerships with "over 3,000 organizations" across "50 plus countries."

Technical Updates Powering AI Mode

Fox provided insights into the product and infrastructure advancements supporting AI Mode and related experiences. A notable achievement was the day-one deployment of Gemini 3 Pro in Search, marking the first time Google launched a "frontier model" directly into Search on its release day.

He also mentioned "generative layouts," where the AI model can dynamically generate user interface code for specific queries. To maintain a fast and responsive experience, Google employs "model routing," directing simpler queries to smaller, faster models while reserving more complex prompts for more powerful, intricate models.

Why This Matters

The continued delay of personalized AI Mode answers, particularly those leveraging opt-in Gmail context, means users currently lack a key promised feature without a clear public timeline. Consequently, users are adapting by providing more explicit context in their queries. If this trend persists, it could influence how publishers structure content, potentially favoring longer, more situation-specific answers to meet evolving user search behaviors.

Looking Ahead

Despite being in its early stages, AI Mode's impressive 75 million daily active users highlight its growing influence. This substantial user base makes it a critical area for publishers and SEO specialists to monitor for visibility and evolving search dynamics.