Google privacy policy change could allow Gmail and Photos data in ad targeting

What happened

Google has updated its privacy policy so that personal data from Gmail, Google Drive and Google Photos could be used to target ads for people who use its personalised AI features. The change was first spotted by Android Authority and reported by The Verge.

The revision adds an exception to a commitment Google has made about not using content from those three services for personalised advertising. Under the earlier wording, that assurance was presented without qualification. The updated text ties the exception to personalised AI functionality, so the position is now conditional rather than absolute.

The reporting available so far centres on the policy wording. It does not set out a start date for any ad changes, the ad formats involved, or how many users the exception could affect.

Why it matters

The story sits at the intersection of AI personalisation and data governance. Useful AI assistants depend on access to context such as emails, documents and photos, and that dependency is what makes the experience feel tailored. The same access also raises the question of how far a data-use promise extends once a company adds a new capability on top of existing products.

For leaders in experience and digital transformation, the lesson is about trust as much as compliance. When a service changes the terms of a promise people relied on, the change is felt as an experience event, not a legal footnote. Organisations building AI features into established products will need to explain clearly what each feature draws on, and when.

The Renascence take

Most readers will focus on the advertising angle, but the more important detail is the structure of the change: a clear promise followed by a conditional exception. Behavioural research consistently shows that people anchor on the headline assurance and rarely revisit the fine print, so the exception does most of its work quietly. That gap between what people believe they have agreed to and what the policy now allows is where trust is lost.

Treat every change to a data-use promise as a service moment, not a legal update. If an AI feature needs access to personal content, make that access a visible, opt-in decision at the point of activation, written in plain language. Operators who explain the trade-off up front will keep the trust that quieter changes put at risk.