Training AI is Fair Use, Microsoft-OpenAI Rivalry Heats Up, Meta Poaches More AI Talent, OpenAI's Hardware Plans, & Apple Eyes Perplexity
AI training is "fair use", but piracy could hinder competition, Microsoft and OpenAI compete, Meta hires more AI talent, OpenAI's hardware plans revealed, and Apple eyes Perplexity
Training AI is “Fair Use”, But Piracy Could Hinder Competition
A U.S. federal judge ruled that Anthropic’s use of copyrighted books to train their Claude chatbot qualifies as “fair use” due to its transformative nature, which is a major win for AI developers.
Judges use 4 key factors for resolving fair use disputes, which are guidelines that courts adapt on a case‑by‑case basis:
The purpose and character of your use
The nature of the copyrighted work
The amount and substantiality of the portion taken, and
The effect of the use upon the potential market.
Anthropic passed all 4 factors. However, part of the case will go to trial, which is that some of the books Anthropic scanned were pirated.
In other words, Judge Alsup ruled that you may set up a major operation of buying physical copies of books (which are used, meaning the copyright holder isn’t benefiting), tearing them up, and then scanning them. However, obtaining digital versions of the books is not fair use.
Big tech has the resources to buy the books, tear them up, and then scan them. US-based startups may not have that luxury, and companies outside of the USA will not be held by these restrictions.
So that’s the open question. Do we want this barrier in place, which will limit US competition in AI? That’s a question ultimately for Congress, regardless of how the trial goes.
Microsoft-OpenAI Rivalry Heats Up in Enterprise AI Market
OpenAI is capturing more market share from their major investor, Microsoft, as both are competing in the enterprise AI market.
Despite Microsoft's $14 billion investment in OpenAI, Microsoft's sellers are struggling to differentiate their integrated OpenAI solution Copilot from the widely recognized ChatGPT.
In fact, Amgen recently transitioned their enterprise AI tool from Microsoft’s Copilot to OpenAI’s ChatGPT.
Aside from ChatGPT’s wider brand recognition, Microsoft is slower integrating new OpenAI models compared to ChatGPT.
Microsoft’s workplace AI CMO Jared Spataro defends this approach, stressing the need for reliability in enterprise settings. However, as clients like Amgen test both platforms, employees often perceive ChatGPT as more cutting-edge, highlighting Microsoft’s challenge in balancing security with AI’s rapid innovation.
Meta Aggressively Pursues More AI Talent
Meta is buying as much AI talent as possible after their disappointing Llama 4 release. After recently acqui-hiring Scale AI CEO Alexandr Wang for $14.3 billion, Meta hired Safe Superintelligence’s CEO Daniel Gross and former GitHub CEO Nat Friedman.
They have also met with AI startups like Runway and Perplexity AI.
On the one hand, hiring the best and the brightest in AI almost seems like an admission that Mark Zuckerberg needs help driving their AI business.
And while the money he’s spending seems big, Meta is a $1.8 trillion company, so they haven’t even spent 1% of their enterprise value. With $67 billion annual income, Meta will make the money they have spent in less than 5 months.
So Zuckerberg is by no means betting the company for the brightest minds in AI. And if they become a leading AI company, their market cap will likely increase far more than 1%.
If you want a deep dive about why Meta can capitalize so much from generative AI, let me know in the comments.
OpenAI’s Hardware Project Revealed from Court Documents
OpenAI’s hardware project with former Apple designer Jony Ive is progressing, with court documents clarifying the product won’t be a wearable or in-ear device, and is unlikely to launch before 2026.
Revealed through a trademark lawsuit from audio startup Iyo, the lawsuit alleges OpenAI’s use of the name “io” infringes Iyo’s brand. This forced OpenAI to remove public references to “io,” which they couldn’t be clearer about in their io launch video.
This legal challenge seems minimal, but underscores a key difference in resources that Jony Ive and his team must adapt to at OpenAI.
Apple was the 2nd most profitable public company in the world (behind Saudi Aramco), and had the resources to tackle any legal challenge. They could also flex their muscle with manufacturers like Foxconn, given that they sell over 200 million iPhones each year.
Things are different at OpenAI. Cash is already being burned training their AI models, and well funded competitors will slow them down as much as possible. On top of that, they don’t have the same manufacturing leverage as Apple, as they haven’t sold a single hardware device. They’ll need to be far more accommodating to manufacturers, which was never a consideration at Apple.
Tech is constantly evolving, we’ll see how Jony Ive and his team evolve to OpenAI’s new paradigm.
Apple Eyes Perplexity AI Acquisition to Boost Siri
Apple is exploring a potential acquisition of Perplexity AI, valued at $14 billion, to enhance their struggling Siri and “Apple Intelligence” initiatives.
Perplexity’s conversational AI search engine could address Apple’s weaknesses in search and conversational interfaces, especially as its Google partnership faces antitrust scrutiny.
This move signals Apple’s urgency to catch up in the AI race, where it lags behind OpenAI and Google. Acquiring Perplexity would accelerate Apple’s AI capabilities, namely in search, integrating a proven solution into Siri and Apple Intelligence.
However, Apple has a lot to gain with their OpenAI partnership. As Google is the vertically integrated player in AI, partnering with OpenAI seems very beneficial for both sides. ChatGPT has over one billion users, and there are over 1.4 billion iPhones worldwide.
But even if Apple goes ahead with Perplexity, OpenAI might as well partner with Apple. Google is already vertically integrated with Gemini, so who else would OpenAI partner with?