OpenAI is pitching a new deal for its data centers, Apple is finally bringing AI to iPhones in China, and publishers are getting one shared line to say no to AI training.
Today’s episode of Daily Tech and AI sounds, at first, like three unrelated stories. Look closer and they share a single question: who gets a say before AI moves in? OpenAI wants public consent for the buildings that power its models. Apple has had to negotiate consent, and a local partner, to ship its AI in China at all. And publishers are gaining a simpler way to tell AI companies to leave their work out of training. Naming, access, and consent are no longer side issues for AI. They are the negotiation itself.
OpenAI’s Post-Midterm Push: “AI Compute Campuses”
OpenAI is preparing a policy and public relations push for the morning after Tuesday’s midterm elections, according to reporting from Axios picked up by TechCrunch on November 1. Chief global affairs officer Chris Lehane told Axios that the midterms will be “a bit of a movie trailer” for 2028, when he expects AI to be the issue in the presidential election. This is not just about one election cycle; it is about setting the terms for the next two years.
The most talked-about element is a name. OpenAI may start calling its data centers “AI compute campuses.” The idea, as Axios describes it, is that these sites are not just warehouses full of servers. They are long-term investments in computing power, energy infrastructure, and skilled jobs, and OpenAI wants the name to carry that argument. A new name does not change the power bill, but it signals how OpenAI wants these projects to be judged.
Community Compacts and a Say for Neighbors
The part worth watching is what OpenAI plans to offer the people who live near these buildings. The company plans to offer community compacts with the towns and counties that host its facilities, and to propose state legislation that gives residents a meaningful say in how projects are built and what benefits they receive. Axios points to a project in Pike County, Ohio, as the model: a site with room for up to eight gigawatts of capacity that OpenAI will lease for twenty years from SoftBank’s SB Energy, with Nvidia guaranteeing part of the lease.
That context matters because data centers have become a campaign issue in their own right. Axios reported that $173 million was spent on general election advertising about data centers, almost all of it negative. OpenAI’s push also reaches further: mandatory safeguards for the most powerful AI systems, which it is calling “speed limits” so humans stay in control; showing people AI helping them in daily life; and a U.S. role in developing global standards. Lehane described the debate around data centers as “good and healthy” — a notable tone from a company that needs to keep building.
Precision matters here. This is a reported plan, not an announcement. OpenAI has not published a formal definition of an “AI compute campus,” and the push is timed for the lame-duck session of Congress, which opens November 9, with new members sworn in January 3. The name may or may not stick. But the timing tells you OpenAI now treats public consent as a constraint on building, not just a communications problem.
Apple Intelligence Finally Arrives in China
Apple Intelligence is finally coming to China. Apple announced on its Chinese newsroom site that the first Apple Intelligence features will reach iPhones in mainland China, with testing in the latest iOS developer beta starting November 4 and the full release coming with iOS 27.3 in December.
This has been a long wait. iPhones sold in China have shipped without Apple Intelligence since the features first appeared in other markets, while phones from local rivals arrived with assistants of their own. Apple Intelligence was approved in China in July, which signaled that a launch was getting near, and Reuters reported in August that Apple had trained its own language model specifically for China.
What China Gets — and What It Doesn’t
The first wave brings the earlier set of features, not the newest ones:
- Notification summaries
- Real-time translation
- Rewriting and proofreading tools
- Visual intelligence
- Removing objects from photos
What makes the China version different is the local partner. Apple says it has worked with Alibaba to add support for its Qwen chatbot in some experiences, similar to the way iOS integrated ChatGPT in other markets back in 2024. What is missing is the new Siri AI — the rebuilt assistant Apple has been rolling out in other markets stays out of the China release for now. So this is a first step, not the full Apple Intelligence experience, and the December release will show how clearly Apple explains the privacy choices to users when a question goes to Qwen.
One Shared Line to Opt Out of AI Training
The third story may be the most consequential, and it turns on a single line of text. On October 30, OpenAI, Google, Microsoft, and Anthropic said they will honor one shared signal in robots.txt, the plain-text file that tells crawlers which parts of a site they may visit. Instead of writing a separate rule for every company’s crawler, a site owner can write one group stating its preference against generative AI training, and every participating provider is supposed to read the same preference.
The proposal comes from an Internet Draft written by Kevin Kelley, a technical lead for standards at Anthropic, for the Internet Engineering Task Force’s AI Preferences working group. It defines a shared policy token for AI training. A site adds a group under that token, then uses the familiar allow and disallow lines underneath it: disallow everything to opt the whole site out of training, or allow just one folder — like a press kit — while keeping the rest out.
The Catch Is Timing
These companies are not all switching it on at once, and the gap is the most practical fact in the announcement. Anthropic has already begun implementing support. OpenAI says it aims to implement the signal in the coming months, while keeping its existing crawler controls in force. Microsoft will start honoring the new preferences on February 1, 2027, as its Bing crawler revisits pages. Google aims to support the token in 2027, and says that where a site opts out with the shared token and does not separately mention Google-Extended, it will treat that as blocking Google-Extended, its own AI training control.
The limits matter just as much. The signal covers training only. It is not a directive for crawling, it does not remove a site from search results, and it does not pull back content already collected or trained on. It is a preference, not a lock: scrapers that ignore robots.txt will ignore this too. And it is not yet a formal standard — the draft is an individual submission to the Internet Engineering Task Force and expires in May 2027 unless renewed or replaced. Meta, for its part, has not confirmed it will honor the token, though its communications head called creating a common signal an important step. For now, the shared line reaches four of the biggest Western AI labs, not the whole industry.
The Thread That Ties Them Together
Put the three stories together and the pattern is clear. OpenAI wants a new name and a new deal with host communities. Apple is finally shipping AI in China, but on China’s terms, with a local partner and without its newest assistant. And publishers are getting one shared sentence to say no to AI training, even if that sentence only works as fast as each company turns it on. This week, that negotiation is happening in Congress, in China’s app stores, and in a text file at the root of every website. Tuesday’s votes will show what actually changed — and we will be back tomorrow with what did.
Sources
- OpenAI post-midterm ‘AI compute campuses’ push: TechCrunch
- Apple Intelligence comes to China with Alibaba Qwen: The Verge
- Shared robots.txt AI training opt-out: The Verge
Adapted from the November 2, 2026 episode of Daily Tech and AI.
Daily Tech and AI is AI-voiced and produced by Angelo Ruiz.
Leave a Reply