Apple is negotiating deals worth upwards of $50 million with major news publishers including New York Times, Washington Post, and Wall Street Journal to license articles, data, and access that could train its artificial intelligence services, according to people familiar with the matter.
The Backstory
Artificial intelligence has become an increasingly important focus area for Big Tech companies like Apple. With chatbots like ChatGPT exploding in popularity and demonstrating new heights of AI’s potential, Apple likely feels pressure to keep pace in the AI race.
Up until now, Apple has taken a more privacy-focused approach to AI and machine learning. Rather than training models on large swaths of user data like some other tech giants, Apple has emphasized on-device processing to protect user privacy.
However, generative AI models that can write articles, poems, code and more require massive datasets to train effectively. Licensing quality data from leading news publishers could help Apple rapidly advance its AI capabilities while still protecting privacy.
Details of the Deals
According to sources cited by the New York Times, Wall Street Journal and other publications, Apple’s deals with news publishers would:
- Span multiple years
- Pay out upwards of $50 million total to the publishers
- License access to archive articles, metadata, and other proprietary data
- Allow Apple to train AI models using the licensed content
The agreements are reportedly still under negotiation and final terms may change. But Apple seems committed to striking deals quickly to begin advancing its AI engines.
In return for the content licenses, Apple may promote the news publishers more prominently in Apple News and other services. But specific benefits to publishers are still being determined.
Apple’s AI Goals
Apple has not officially stated plans for AI products or features that these news licenses might empower. But based on the company’s strengths and chatbot leader ChatGPT’s capabilities, some possibilities include:
- Enhanced search – More intelligently answering complex search queries with summarized explanations drawn from licensed news content
- Data analysis – Identifying insights, trends and conclusions across vast news archives
- Content summarization – Automatically creating abridged versions of articles
- Personalized news curation – Suggesting the most relevant headlines based on reader interests
- Conversational assistance – Powering Siri and other chatbots to discuss news events and concepts found in articles
- Writing aids – Helping users craft news stories, essays, reports and other writing using data and linguistic patterns found in the publisher content
These hypothetical AI features could significantly upgrade Apple services like Siri, Spotlight, Apple News and iCloud.
Competing With ChatGPT and Google
ChatGPT’s launch in late 2022 took the world by storm, demonstrating new heights of generative AI’s potential. Within days, users generated over 1 million conversations with the chatbot tool.
Apple likely feels pressure to keep pace with pace-setting AI innovations like ChatGPT. Licensing news publisher content to train its models could help it close the gap with ChatGPT’s capabilities and launch Apple’s own standout AI.
Google also recently announced a chatbot named Bard that directly rival’s ChatGPT. And Google has already indexed vast swaths of content and trained AI models on more data than perhaps any company outside China.
High-quality licensed news content combined with Apple’s privacy-focused approach could be its best shot at standing apart from ChatGPT and Google in the red-hot AI space.
What Publishers Stand to Gain
For news publishers, Apple’s content licenses offer both financial gain and heightened visibility. The multi-million deals provide an influx of revenue that could fund reporting and operations.
Additionally, Apple services like Apple News, Stocks, and Siri reach millions of device users daily. If those services prioritize content from licensed news partners as part of the deal terms, publishers could see significant referral traffic and subscriber boosts.
However, some publishers may still have hesitation around licensing articles. There are concerns that enhanced AI summarization could enable aggregating of content without directly accessing publisher sites. And publically accessible archive content could empower new startup competitors.
What’s Next
According to the New York Times, Apple and publishers hope to finalize terms within the next several weeks. So Apple will likely tap licensed content to immediately accelerate AI development.
If Apple indeed utilizes publisher archives to launch standout AI features, ChatGPT and other generative models have demonstrated the viral-like appetite for this technology. So Apple will need to strategically market the capabilities to device users in order to maximize adoption.
And as AI becomes further cemented as a computing pillar, Apple will need to communicate how its privacy-centric approach stands apart from Big Tech competitors. With smart messaging, Apple could position itself as the leader in both AI and data privacy.
Additionally, after striking deals with leading English publishers, Apple may look to license content from international and niche publishers to further improve linguistic and topical breadth.
While nothing is final yet, Apple’s exploration of major AI licensing deals signals its seriousness around rapid generative AI advancement. And with the world’s most valuable brand applying itself to the space, it could compel competitors to reassess their own strategies.
| Publisher | Estimated Archive Size | Unique Visitors |
|---|---|---|
| New York Times | 5 million articles | 5 million daily |
| Washington Post | 3 million articles | 2 million daily |
| Wall Street Journal | 4 million articles | 1 million daily |
To err is human, but AI does it too. Whilst factual data is used in the production of these articles, the content is written entirely by AI. Double check any facts you intend to rely on with another source.
