OpenAI struck a licensing deal that puts real photography inside ChatGPT. Apple and the car industry, meanwhile, are reshaping the screens you stare at on the commute and at the dinner table.
Three product moves landed within roughly 48 hours of each other in late June 2026, and each one targets a different corner of an ordinary day. None arrived with a flashy keynote. Together they point to where the largest technology companies believe consumer AI is heading next: into the everyday apps and dashboards people already open without thinking.
Here is what changed, and why each shift matters more than the press releases let on.
OpenAI Stops Letting ChatGPT Guess at Pictures
Getty Images has signed a multi-year partnership with OpenAI that places its licensed photo and video libraries directly inside ChatGPT and OpenAI's search results. The practical effect is simple. Ask ChatGPT something that calls for a visual, and you can be served a licensed Getty image with clear provenance instead of an AI-generated approximation or an unsourced picture pulled from the open web.
The deal carries weight because of who Getty used to be in this fight.
In September 2022 the company banned AI-generated art from its library outright. Months later it sued Stability AI for copyright infringement, a case that was thrown out late in 2025. For a stretch, Getty was one of the loudest holdouts against the generative AI industry.
That position softened in stages. Getty built its own generative tool trained on its owned library and ran it on NVIDIA's Edify model, attaching a royalty-free license to every output. Then in October 2025 it signed with the AI search startup Perplexity, on the condition that Perplexity display image credits linking back to the original source. The OpenAI agreement extends that same logic to a far larger audience.
One detail is conspicuously absent. Getty has not said whether OpenAI can use its images to train models. The earlier Perplexity arrangement explicitly forbids training use, so the silence here is the part to watch. Getty chief executive Craig Peters framed the partnership around making AI search "more useful and more trustworthy," a pitch aimed squarely at the credibility gap that unsourced AI images have opened up.
Apple's Real iOS 27 Story Has Nothing to Do With Siri
Siri's long-promised AI overhaul took the headline slot at Apple's Worldwide Developers Conference in June. The more interesting story sits underneath it.
Rather than betting everything on people adopting a smarter Siri, Apple has threaded AI through the apps you already use, and most of it is built to disappear into the background. The features are live in the developer beta now, reach the public beta shortly, and ship broadly with iOS 27 this fall.
Bill splitting is the most relatable example. Photograph a restaurant receipt and Apple Intelligence reads the items, quantities, tip, and total, then lets you fire a request into a group chat where everyone taps the dishes they ordered. It handles half-portions and divides tax and tip on its own. Payment runs through Apple Cash with the same double-click you already use.
Password security gets an agentic upgrade. Apple's new tool scans for weak and breached credentials, then acts without waiting for you: it navigates to the affected websites, signs in, and replaces the compromised passwords with stronger ones, sparing you the manual reset on each site.
Other additions are smaller, but they stack up.
Messages will surface one-tap suggestions tied to what you are actually discussing. A friend asks you to bring something, and the app offers to drop it into Reminders. Someone requests photos from an event, and Apple Intelligence pulls the right shots using its read on locations, dates, and faces. A feature called Call Context displays details you need mid-call, such as an airline confirmation code, by reading your Mail entirely on the device. Calendar events can be created from a plain spoken description. And Shortcuts, long gated behind a learning curve that scared off casual users, now responds to a single sentence, so you can tell your iPhone to text your partner your ETA when you leave work or to switch on the porch light when a DoorDash order pulls up.
The Home app gains a sanity feature that smart-home owners will appreciate. Instead of firing a separate alert when someone opens the garage and then walks inside, it bundles the sequence into one notification describing what actually happened.
The through-line is the point. None of these announce themselves as artificial intelligence. They just make the phone feel more capable.
The Car Industry Is Quietly Killing Android Auto, and Drivers Will Pay for It
For about a decade, a quiet bargain held between drivers and automakers. You bought the car, and the carmaker let you plug in your phone to run Android Auto or Apple CarPlay so your music, maps, and messages came along at no extra cost. That bargain is starting to break.
General Motors, one of the largest automakers on the planet, has pulled Android Auto from its EVs and plans to remove it from the rest of its lineup. In its place GM is building a conversational system powered by Google's Gemini. Rivian and Tesla never offered Android Auto in the first place. Most 2026 models still include it, yet the direction of travel is clear, and the reasons behind it do not flatter the companies making the call.
Data is the first reason. When you run Android Auto, Google collects GPS and mapping data it can sell to advertisers, and none of that reaches the carmaker. GM argues the setup starves it of information it needs, for instance to improve EV charging and to plan routes around battery range and charger availability. GM also carries baggage here: it paid a $12.75 million fine after breaking California privacy law and is now barred from selling driver data to advertisers anyway.
Money is the second reason, and the one drivers should scrutinize.
GM has openly described "subscription revenue opportunities" tied to running its own infotainment stack. The industry's record on this is poor. BMW once tried charging $18 a month for heated seats. Built-in systems also depend on an active cellular connection, since your phone is no longer doing the work, which means a data subscription eventually comes due. GM ships eight years of OnStar service with new vehicles, though what happens afterward goes unstated. Rivian and Tesla each charge $150 a year for premium connectivity. Even Kia, which fully supports Android Auto, parks features like remote locking behind trials that convert to paid plans.
Rivian frames the change as inevitable progress, telling The Verge that deep in-car AI integration makes "the entire CarPlay debate completely obsolete."
The counterweight is the buyers. GM's decision drew immediate backlash, with plenty of people saying flatly that they will not buy a car without Android Auto. A broader revolt against subscriptions of every kind is brewing, and asking drivers to rent features inside a vehicle they already bought tends to land badly. Automakers have also repeatedly proven that they are bad at building their own infotainment software, which is the reason phone projection won the dashboard in the first place.
What These Three Moves Reveal About Big Tech's Next Play
Set the three side by side and a single strategy comes into focus. Each company is pulling the AI experience inside its own walls.
OpenAI is bringing licensed, sourced content into ChatGPT so the chatbot becomes a place people trust for visuals, not only text. Apple is embedding intelligence into Mail, Messages, Calendar, and Photos so the value shows up without anyone opening a separate AI app. GM is yanking the dashboard back from Google so it owns both the data and the recurring revenue that phone projection handed over for a decade.
The pattern rewards the platform owner. Whether it rewards the person holding the phone or sitting behind the wheel is far less settled, and that comes down to details still being worked out: how Getty's images get attributed and whether they feed future training runs, how much of Apple's on-device privacy promise survives contact with the public beta, and how aggressively carmakers price the subscriptions they are clearly planning to sell.