Apple is reportedly trying to secure professional publisher content for generative AI training, offering multi-year licensing deals worth at least $50 million. The talks put a clear price on high-quality written archives, while also exposing a difficult question for media companies: what happens when the content that built their brands becomes training material for AI systems?
Apple is seeking publisher archives for AI training
According to The New York Times, which cited four people familiar with the matter, Apple has reached out to several major publishers and news organizations in recent weeks. The company is said to be looking for licenses that would allow it to use their content to train AI systems.
The reported offers are large by normal content-licensing standards. Apple is said to be proposing multi-year arrangements worth at least $50 million.
Among the companies Apple has approached are Condé Nast, NBC News and IAC. IAC owns People, The Daily Beast and Better Homes and Gardens.
The move reflects a broader reality in generative AI: professional texts are valuable because language models depend on large bodies of written material. For a company trying to build stronger AI products, publisher archives can offer edited, structured and widely read language at scale.
The deal terms raise hard questions
Some publishers are reportedly uneasy about the scope of Apple's proposed licensing terms. Their concern is not only the amount of money on offer, but how much access to publisher archives Apple would receive in exchange.
Another concern is liability. The source article says publishers would be held legally liable for Apple's use of their content. That shifts part of the risk back onto the organizations providing the material, even though Apple would be the company deploying the AI systems.
For publishers, that creates a difficult trade-off. A licensing agreement could bring meaningful revenue and establish permission-based AI training as a business relationship. But broad archive access could also reduce control over how valuable editorial material is used after it leaves the newsroom environment.
- Archive rights: publishers must consider how much historical material they are licensing.
- Legal exposure: the reported terms would make publishers liable for Apple's use of their content.
- Future use: it is not yet known how Apple would apply generative AI in a news context.
The strategic risk is competition
The most sensitive issue is not simply whether Apple can train AI on news content. It is what Apple might eventually do with news-trained generative AI inside its hardware and software ecosystem.
If Apple were to use such systems to take over parts of the news business, publishers could end up competing with products trained on their own work. That is the core tension in the talks: the same content that creates a licensing opportunity could also support tools that weaken publisher relationships with readers.
This risk is especially important because Apple has a broad ecosystem. The source article does not say what Apple plans to launch, and that uncertainty is part of the problem for publishers. Without knowing the final product, it is harder to judge whether a licensing deal is a partnership, a defensive move or a long-term threat.
The article compares the situation with artists whose material has been used to train AI without permission or royalties. The difference here is that Apple is reportedly asking publishers first. That makes the discussion more formal, but it does not remove the business concerns.
Apple's approach differs from other AI deals
Some news industry executives are reportedly optimistic because Apple is seeking permission rather than simply using content without consent. That could make the talks a possible model for AI licensing, even if the terms remain contested.
The source article contrasts Apple's approach with what it says Google or OpenAI do: taking what they need without asking. It also notes that Apple could still find ways to train capable generative text systems without these publisher licenses.
OpenAI has recently announced partnerships with Axel Springer and the Associated Press in the US. In both partnerships, OpenAI receives journalistic content for AI training and provides AI technology that can help optimize production processes. Content from Axel Springer publications will also be distributed in ChatGPT.
Those agreements show how AI companies and news organizations are beginning to test new arrangements. The basic exchange is content access on one side and money, technology or distribution on the other. Apple's reported offers appear to fit into that larger pattern, while still carrying their own legal and strategic questions.
Apple is trying to catch up in AI
The talks also matter because Apple has been less visible in the recent AI wave than OpenAI, Microsoft and Google. The source article also names startups such as Anthropic and Midjourney, along with a vibrant open-source scene, as major parts of the current AI landscape.
Apple's absence has started to change. The company is working on AI tools, including a large language model framework called Ajax and a chatbot service called AppleGPT. The project has become an important part of the company.
According to Bloomberg, Apple is already using the language model internally. It is designed to help employees build prototypes, summarize text and answer questions about the data it has been trained on.
Apple's language modeling team is said to be small, about 16 people, but the budget is large. Millions of dollars are said to go into AI training every day. The "Foundational Models" team is led by AI engineer John Giannandrea, whom Apple poached from Google in the spring of 2018 and promoted to the executive team the following winter.
Apple has also recently released MLX, an efficient machine-learning framework for Apple chips. With MLX, users can locally train or fine-tune a Transformer language model, generate text with Mistral, create images with Stable Diffusion and perform speech recognition with Whisper, among other features.
Taken together, the reported publisher talks and Apple's internal AI work point in the same direction. Apple is building more serious generative AI capabilities, and high-quality publisher content may become one of the resources it wants most.