LibreOffice Says No AI Is a Feature, Not a Bug

The AI industry is having an identity crisis about what "more" even means anymore, and one software project just decided the answer is "less."

LibreOffice Plants a Flag: No AI Is the Feature

LibreOffice, the free, open-source alternative to Microsoft Office used by millions of people and organizations worldwide, announced it has no plans to add AI to its default setup. The reason is simple: privacy. When you use AI features in most software, your documents are typically sent to a company's servers to be processed, which means your content leaves your computer. LibreOffice is saying it won't do that by default — they're calling it a feature instead of an excuse. For lawyers, HR professionals, teachers, or anyone working with sensitive documents, that changes the conversation from "why doesn't it have AI?" to "why does yours send my files to a server I don't control?" Watch for legal software vendors to start doing this next — they have the most to lose if a client's privileged documents touch an outside server.

https://techcrunch.com/2026/10/06/libreoffice-says-no-ai-is-now-a-software-feature/

Mistral Drops "Le Chonk," a One-Trillion-Parameter Model

French AI lab Mistral released Mistral Large 4, internally nicknamed "Le Chonk," a model with one trillion parameters. Parameters are roughly the learned settings inside an AI model that determine how it thinks and responds. More usually means smarter, but a smaller model trained on better data can beat a bigger one — Mistral's whole reputation was built on proving that. What's notable here is that Mistral is releasing this as "open-weight," meaning other companies and developers can download and run it themselves without paying Mistral. That's rare at this scale: the closed models competing at this level — GPT-4o, Gemini Ultra, Claude Opus — aren't available to self-host at any price. Running a one-trillion-parameter model requires serious hardware, but for large enterprises or governments with the infrastructure, the option to keep everything in-house just got meaningfully more capable.

https://techcrunch.com/2026/10/06/mistrals-new-1t-model-aims-to-leapfrog-closed-and-open-rivals/

Google Starts Charging for Models It Currently Gives Away Free

Starting October 9th, Google is pulling Gemini Flash and Gemini Pro from its free tier. Free users will only get Flash Lite, which is a lighter, less capable version of the model. To access the standard Flash model, you will need to pay $4.99 per month for Google AI Plus. This is a pattern, not an accident. Labs spent the last two years giving away capable AI for free to build habits and displace competitors, and now they are converting those habits into revenue. If you have built any personal or team workflow around free Gemini access, check which model you have actually been using — the cutover is October 9th.

https://www.theverge.com/ai-artificial-intelligence/1005451/google-gemini-free-flash-lite-only

Pinterest Turns Saved Looks Into Salon Appointments

Pinterest rolled out Beauty Guides, an AI feature that takes hair and nail images you have saved and translates them into specific, actionable information: the correct salon terminology to ask for, estimated cost ranges, how long the appointment will take, and how much upkeep is involved. You screenshot a balayage from someone's Instagram and then have no idea what to call it at the salon — that's the problem this solves. It's also a smart business move for Pinterest, whose entire value proposition has always been aspiration without friction. The more interesting signal is the model underneath it: there are dozens of industries where customers show up knowing what they want but not how to ask for it — interior design, auto repair, physical therapy, custom tailoring. The businesses that figure out how to close that gap with AI before their competitors do will have a real advantage.

https://techcrunch.com/2026/10/06/pinterests-ai-now-turns-beauty-pins-into-action-plans/

The real question today is not which AI is most powerful. It is which direction your organization is moving: toward tighter integration with AI services you do not control, or toward tools that keep your data where you can see it.

← More daily briefs