OpenAI Picks An Apple Fight
Guys. It’s the last day of your big, highly-publicized and insanely-scrutinized trial against your former co-founder — who just happens to be the richest person in the world. Can you just, I don’t know, wait a while until you get into another high profile fight? Maybe a year? How about a month? Can we at least get a week breather? A few hours?
At this point, I think we have to recognize that OpenAI is going to be one of those companies that is simply always at war. They won’t need a “peacetime CEO” because there will never be peace. When one battle ends, the next begins. Elon Musk, the said co-founder. Anthropic, started by former OpenAI folks, of course. Microsoft, long the main benefactor. I mean, the entire company was basically founded to combat Google.
And now Apple.
Yes, OpenAI is at least thinking about legal action against the iPhone maker, reports Bloomberg. And The Financial Times. And The Information. And The New York Times. And undoubtedly others. Because it sure seems like OpenAI sources are willing to talk to anyone and everyone about such prospects. It's almost like they held an anonymous press conference on the matter. The sources include at least one OpenAI "executive" – albeit an unnamed one. Clearly, they are trying to send a message.
This is so wild that you almost have to wonder: has anyone at OpenAI ever met anyone at Apple before? This method of communication is not going to go over well in Cupertino. Why not just hurl a flaming bag of dog shit on to the steps of 1 Infinite Loop?
Here's Mark Gurman:1
Apple Inc.’s two-year-old partnership with OpenAI has become strained, according to people familiar with the matter, with the AI startup failing to see the expected benefits from the deal and now preparing possible legal action.
OpenAI lawyers are actively working with an outside legal firm on a range of options that could be formally executed in the near future, said the people, who asked not to be identified because the deliberations are private.
That could include sending the iPhone maker a notice alleging breach of contract without necessarily filing a full lawsuit at the outset, according to the people. OpenAI enlisted the outside firm in recent days to help with the situation.
Those deliberations are no longer private. So OpenAI better hope any contract they have in place with Apple is pretty ironclad. But the notion that they're merely thinking about sending Apple a notice about a breach of contract – and the fact that they're using this public negotiation maneuver – suggests that it may not be.
Per the reports, it sounds like OpenAI has been trying to rework their deal with Apple in recent months, but clearly such talks have gone nowhere. So they could sue, or threaten to sue, but floating that idea out there ahead of actually doing either... well, isn't the best look. OpenAI likely thinks appealing to the public – and perhaps politicians?2 – will help them here. Again, have they met Apple?
OpenAI believed that the companies’ partnership, which wove ChatGPT into Apple software, would coax more users into subscribing to the chatbot. It also expected deeper integration across more Apple apps and prime placement within the Siri assistant.
Instead, Apple’s use of OpenAI technology across its operating systems remains limited, and features can be hard to find.
“We have done everything from a product perspective,” said an OpenAI executive who asked not to be identified. “They have not, and worse, they haven’t even made an honest effort.”
I mean, can you sue someone for a deal not living up to expectations?3 I'm going to go ahead and guess that Apple's lawyers would not have agreed to terms that could possibly put them in such a position.
The real problem is that ChatGPT integration within Siri did ship and... it just sort of fell flat? That was certainly more on Apple because it was buried inside of Apple's famously problematic AI efforts. But I'm just not sure you can literally litigate such product decisions and failures. Was OpenAI promised very specific placement they didn't get? Were they promised some sort of minimum usage guarantees? There famously was no money explicitly changing hands in the deal, so was there some sort of revenue share guidance that was written down at some point? Again, hard to see how Apple would paper any of that.
The reality is that it sounds less like an Apple problem and more like one of sour grapes. Did the company flub their AI efforts to start? For sure. Did that hurt OpenAI? Perhaps – if AI on the iPhone was awesome, more usage would probably have equated to more ChatGPT usage. But we obviously can't know that for sure. I'm sympathetic to the expectations that OpenAI clearly had here (and perhaps Apple even shared) that this deal could have been a massive moment for ChatGPT, certainly for the subscription side, because I thought it would be too! It just didn't play out that way. And so perhaps OpenAI is trying to latch on to the lawsuit Apple just settled around advertising AI features that never shipped?
It's just far more likely that the partnership failed to live up to some loose and vague expectations – on both sides. Sure, that's undoubtedly more on Apple. They dashed OpenAI's hopes and dreams of a Google Search-like deal. But can I sue HBO over the final season of Game of Thrones?
It seems pretty clear that this is all timed to get ahead of WWDC in a few weeks. There, Apple is expected to formally unveil their new AI work, notably the new Siri, now powered by models distilled from Google's Gemini. And a range of new AI options, including integrations with Anthropic's Claude and perhaps any other AI service a user chooses to download. With the former, OpenAI would really like us to know they had no interested in that level of partnership:
OpenAI wasn’t interested in working with Apple on the new models because it felt burned by the initial relationship, according to the people. “Apple has so much market power that they can dictate terms,” the executive said. “We already took this leap of faith with you, and it didn’t work out well.”
Which is a weird statement because it reads more like they didn't like Apple's terms for such an arrangement more so than they were upset about the previous collaboration.
Then again, per earlier reporting (also from Gurman), Anthropic may have been Apple's preferred partner for their AI work after an internal "bake off", but terms were also the sticking point there. It's not clear if OpenAI was in second or third place in that scenario. They'd clearly like us to believe they were in second (or first!) but Apple would sure like us to believe that Google was in first, so...
As for the other app-based integration:
The OpenAI executive said that Apple’s embrace of other AI providers isn’t driving the company’s legal action since the partnership wasn’t meant to be exclusive from the start.
Which is also a funny statement because it reads like they're cool with that because they have to be, legally. That lack of exclusivity is undoubtedly something Apple papered!
To be fair, it does seem like the new "extension"-based integration for third-party AI providers could end up being better than the current situation, simply because it sounds like functionality and discovery will be less buried. And that means OpenAI could ultimately get out of Apple what they thought they were going to in their original deal. But yes, they'll have to battle Anthropic and others now.
Just for fun, let's go ahead and layer in a few more things.
First, Apple probably doesn't love the fact that OpenAI acquired the Jony Ive-founded io with the notion of building products that would break the chains that the iPhone has placed upon us. But they really can't love the fact that the team keeps poaching key Apple talent to help with such efforts.
And what if some of those people are also working on an actual iPhone competitor in the form of a phone, or a smartphone-like device?
I'm not saying that's the reason Apple may not have been as receptive to reworking any deal with OpenAI, but I'm saying that if there were such discussions, that dynamic probably didn't help matters!
As an aside: it seems a bit strange that Apple would even consider reworking their OpenAI deal when they must have known they were about to change the entire AI system, not only with Gemini, but with the aforementioned "extensions". Why bother? Unless OpenAI wanted more preferred placement for ChatGPT – which per above, they're saying they didn't. So perhaps that's the simple reason why Apple and OpenAI's discussions to "rework" their deal went nowhere?
Let's also remember that after the initial OpenAI agreement for iOS integration, Apple was said to be in talks to invest in a round of funding – which would have made sense – and potentially get a board observer role at the company! Microsoft may or may not have torpedoed that (and killed off their own such role in the process). And ongoing OpenAI turmoil at the time may or may not have killed off the investment itself. But again, the fact that this (allegedly) played out after that original agreement probably started the two sides down a rockier path than the one they set out upon.4
Two years ago, I was wondering if we might see Sam Altman (back!) on stage at WWDC to announce their partnership. It didn't happen, but ChatGPT certainly got prime placement there, if not exactly within iOS itself. This year, with this news, I think it's a question of if we'll hear anything about ChatGPT – still the leading consumer AI product and partner of Apple, after all – at all.
Jury deliberations start on Monday for the Musk/OpenAI trial. Hopefully they can wrap in time just in case OpenAI needs to start a new trial.5 One battle after another, indeed.
1 Leaking such news to Gurman hardly seems like an accident given his status as the preeminent Apple reporter. OpenAI is clearly aiming for the Apple audience while sending a message to the company. Again, I'm just not sure how wise of a strategy this is. Optically, it's fun though! Were the TBPN guys consulted on the strategy? ↩
2 One quote given to the FT stands out in this regard: "They are focused solely on extracting a tax for their market position" Someone alert Elizabeth Warren, pronto! ↩
3 The NYT's Mike Isaac smartly recalls the Square/Starbucks deal from over a decade ago that seemed like a massive win for the then-startup but ended up as a sort of disaster (and was killed in their IPO filing, no less!). As Isaac notes, no lawsuits were filed there... ↩
4 There's also some notion that Apple is not happy with OpenAI's privacy standards? Which was an aside in the Bloomberg piece, but obviously has long been a concern with such a deal... ↩
5 Perhaps Judge Yvonne Gonzalez Rogers can oversee this one too, since she's now very familiar with both companies... ↩