In the past week, Google quietly changed a setting that affects almost every business that depends on search marketing: your interactions with Google’s “Search services” can now include saved media (images, files, audio, and video) that may be used to develop and improve Google’s AI models unless users opt out. The immediate takeaway for marketers isn’t panic—it’s preparedness: privacy expectations are rising, regulators are watching, and customers will demand clarity about how their data powers AI.

TechCrunch reports the change introduced new controls called Search Services History and Personalized Recommendations, and explicitly calls out media types like “images, files, and audio and video recordings” as part of what can be stored and used to improve AI systems. For business owners, this is not just a consumer privacy story. It’s a signal that search is moving faster toward AI-first experiences, and data policies will keep evolving alongside it.

What changed: “Search Services History” expands beyond text queries

The notable shift is that Google’s Search-related history isn’t just the text you type. According to TechCrunch, the updated settings can include saved media and recordings that help “develop and improve Google services and technologies, including AI models and safety measures.” That matters because modern search is increasingly multimodal: people search with photos (Lens), voice (Search Live), and other inputs that look more like “content” than a classic keyword query.

TechCrunch also notes that this applies beyond Google Search to other “search services” including Maps, Shopping, Flights, Hotels, Translate, and News. In other words, the behavior that fuels discovery and conversion—local intent on Maps, product research in Shopping, and travel planning in Flights/Hotels—now intersects more directly with AI training and AI personalization.

Why this is an AI marketing story (not just privacy)

As AI-first search expands, marketers will increasingly compete for visibility inside AI-generated answers, not just blue links. That makes user trust and consent a performance variable. When customers believe a platform is over-collecting, they disable features, opt out of personalization, or shift behavior to other channels. The result: less reliable targeting signals and noisier attribution.

There’s also a brand risk angle. If your team uses Google tools in customer-facing workflows—for example, encouraging customers to upload photos for support, quotes, or product matching—you need to know what the platform may retain and how to communicate that clearly. The more multimodal your funnel becomes, the more your marketing operations and compliance operations overlap.

What businesses should do now (practical steps)

  • Audit customer-facing moments where Google tools are used. If you ask customers to upload images (Lens-like workflows), use voice input, or share files, document the tools and flows.
  • Update your privacy and consent language. If your process relies on Google services, add plain-English disclosure about what may be stored and for how long.
  • Train your team on the new knobs. TechCrunch reports users can uncheck “Save Media” separately from the broader Search Services History setting, and can set auto-deletion windows like 3 months, 18 months, or 36 months.
  • Prepare for more AI-personalized SERPs. As personalization expands, segment your reporting by device, geography, and query class to detect when AI features are suppressing clicks or changing the journey.

How this connects to GEO (Generative Engine Optimization)

AI discovery is not happening in just one place anymore. Previsible’s 2026 State of AI Discovery report (covering millions of AI-driven sessions) argues that AI discovery inside Google—through AI Overviews and AI Mode—still represents more AI-influenced traffic than standalone assistants combined, while ChatGPT leads among standalone LLM referral traffic. For marketing leaders, the strategy implication is clear: you must win visibility inside Google’s AI surfaces while also building a presence in assistant ecosystems.

The best GEO playbooks in 2026 look less like “hack the algorithm” and more like “become the most citable source.” That means publishing original evidence, using clear site architecture, and earning third-party authority signals that AI systems trust. But it also means respecting user consent, because trust is becoming a constraint on the data that powers personalization.

CEO takeaway: treat privacy shifts as leading indicators

When platforms change defaults and data policies, it’s rarely isolated. It usually signals a product roadmap shift. In this case, it aligns with an AI-first search trajectory where richer inputs (images, audio, files) become normal. The businesses that win will be the ones that adapt their marketing operations, measurement, and messaging before the next change lands.

If you want help aligning your SEO, GEO, and content strategy to the reality of AI-first search—and doing it in a way that protects customer trust—Real Internet Sales can help. Call 803-708-5514 or visit realinternetsales.com to talk through your next 90-day plan.

Sources: TechCrunch (July 6, 2026) and Previsible’s 2026 State of AI Discovery report coverage.