🧠 Google’s ‘Personal Intelligence’ Will Use Your Own Data

Google is giving Gemini a more personal touch. A new feature called Personal Intelligence lets the app tap into things like your Gmail, Photos, and files so it can answer questions with real context from your own digital life. Let’s dive in!

In today’s insights:

LATEST DEVELOPMENTS

Evolving AI: Google is testing a Gemini feature that links your apps and files to answer with personal context.

Google has launched Personal Intelligence inside the Gemini app, letting users connect data from services like Gmail and Google Photos. The feature lets Gemini reason across personal files, messages, and media to answer questions with richer context. Google says the system does not train directly on inboxes or photo libraries, and remains optional. Early access is limited to paid users, with plans to extend it into Google Search’s AI Mode later on.

Gemini actively uses your own connected data to link signals across apps and deliver answers without manual prompts. Google positions this as on-demand use of personal context, with controls in place to decide when and how that data is accessed.

Personal Intelligence is Google betting that the next AI battleground is your context, not model bragging rights. If Gemini can pull the right detail from Gmail, Photos, Search, and YouTube in one go, everyday jobs like trip planning, picking gifts, finding files, or recalling commitments get faster without jumping between apps. Google keeping it off by default and claiming no direct training on inboxes or photo libraries shows how much privacy trust will decide adoption, right as Apple leans on Gemini for Siri.

AI gets smarter when your input is complete. Wispr Flow helps you think out loud and capture full context by voice, then turns that speech into a clean, structured prompt you can paste into ChatGPT, Claude, or any assistant. No more chopping up thoughts into typed paragraphs. Preserve constraints, examples, edge cases, and tone by speaking them once. The result is faster iteration, more precise outputs, and less time re-prompting. Try Wispr Flow for AI or see a 30-second demo.

Start flowing free

Evolving AI: Alibaba’s Qwen app now completes real tasks like food orders and travel bookings.

Alibaba has rolled out a major upgrade to its Qwen AI app, shifting it from a chat tool into an action-first assistant. Users in China can place instant commerce orders, book trips, and complete payments without switching apps. The update connects Taobao, Alipay, Fliggy, Amap, and other Alibaba services into one interface. A Task Assistant, now in invite-only beta, can make phone calls, process up to 100 documents, and build multi-stop travel plans. The app has already crossed 100 million monthly active users since its beta launch.

When a chat app can pay and book without kicking you out to other apps, it stops being a chatbot and starts being your interface to daily life. Alibaba wiring Qwen directly into Taobao, Fliggy, Amap, and in-chat Alipay payments turns “ask” into “do” in one flow, and the Task Assistant going as far as calling restaurants hints at where this is headed: fewer taps, fewer forms, more delegated errands. With Qwen already past 100M monthly active users since its November beta, Alibaba gets a real distribution edge to push agent-style commerce at scale, right as OpenAI and others are building similar “do the task” agents in the West.

Evolving AI: Wikipedia marks 25 years by signing paid AI data deals with Big Tech platforms.

Wikipedia turned 25 by formalizing paid AI data deals with major model builders. AI firms now pay for high-volume, high-speed access to Wikipedia’s content via its enterprise platform. Human visits dropped 8% as AI summaries replace links, yet bot traffic surged, stressing infrastructure. Wikimedia frames this as cost sharing, not restriction, while keeping Wikipedia free for public use.

When chatbots and search summaries answer the question right on the spot, sites like Wikipedia lose the click but still eat the server bill from nonstop AI scraping. These new Wikimedia Enterprise deals with Microsoft, Meta, Amazon, Perplexity, and Mistral turn that imbalance into a pay-for-access lane, so model builders get clean, structured data at the scale they want, and Wikipedia gets funding that doesn’t come out of donor pockets. If this holds, expect more publishers to copy the playbook: stop fighting bots one by one, sell a sanctioned pipeline, and reinvest in quality and tooling for editors and search.

“When I first opened Attio, I instantly got the feeling this was the next generation of CRM.”
— Margaret Shen, Head of GTM at Modal

Attio is the AI-native CRM for modern teams. With automatic enrichment, call intelligence, AI agents, flexible workflows and more, Attio works for any business and only takes minutes to set up.

Join industry leaders like Granola, Taskrabbit, Flatfile and more.

👉Start for free today

QUICK HITS

🤝 OpenAI partners with Cerebras.

🖨️ Generative AI tool helps 3D print personal items that sustain daily use.

🔞 California launches investigation into xAI and Grok over sexualized AI images.

🗣️ Google’s TranslateGemma: A new suite of open translation models.

 *partner link