“privacy protections”

Who knows what it is? Apple just put out the most ominous Apple PR statement of the AI era and left out all of the concrete details of what it indends to do. Now every AI company with a market cap over $100 million is antitrust lawyering up. What a mess.

That’s an interesting one. So, inside a game made by a random developer, there are multiple ad engines from multiple vendors and they’re monitoring and calling each others APIs while litigating whether it’s appropriate to do so.

The user should be free to decide what to give their AI assistants access to. Apple’s operating systems team should support user choice. Having Apple be a non-fiduciary intermediary so it can block Apple competitors and extort junk fees will destroy Mac as an open platform.

Apple did not ship a privacy feature. It shipped a veto on Meta’s agent. Muse’s whole pitch is that it can go do the thing: open the app, read the thread, book the table. The moment that “thing” is your iMessage history, Apple’s permission model and Meta’s product are the same fight. This is the agent era’s version of the App Store tax. The model company wants the action. The platform company owns the surface the action runs on. Whoever controls the permission prompt controls the agent.

This is wildly anticompetitive since Apple also blocks all competing web browser engines, so it’s impossible for iOS users to choose a genuinely different browser. Cut off the supply chain for news sites’ remaining ad revenue and then…? Launch an Apple web ad network?

The wording of Apple’s statement demonstrates that their PR team was behind the astroturfy campaign to attack tools interoperating with iMessage many weeks ago. It’s that “you might have the computer owner’s permission but not other conversation participants” line.

“A few extra steps” landed Apple in contempt of court in Epic v Apple, because they can design a “few extra steps” to make all of their competition economically unviable while pumping up their own inferior products.

Siri’s Law: You don’t have to use Siri. You just can’t use anything else.

Users should be free to decide. If Apple’s permission screen is unclear, I support Apple making it clearer, but blocking entire categories of competition to their own awful AI assistant is massive theft of rights that rightly belong to Mac buyers.

Adobe went quietly in the night. OpenAI, Anthropic, XAI, and Meta will not. Nor will antitrust enforcers on at least 3 continents.

Is it straightforward? Apple put out an ominous yet super vague statement. Since they could have said exactly what they were going to do, and chose not to, either they’ve forgotten how to do PR or their plans are truly ominous and will spark massive legal fights.

Apple is going to make it even harder to productively use macOS in the age of agents? Bold move. Let's see how it plays out!

Very explicit as in a screen saying “Here is very explicitly what this permission grants… Accept/Cancel” or very explicit as in the abomination of 15 scare screens, traps and dead ends as Apple unlawfully launched on iOS in EU to thwart competition? epicgames.com/site/news/appl…

AI users: “MacOS permissions suck. They get in the wa-” Apple (interrupting): “Understood! We’ll make sure you get way more popups and notifications about permissions!”

Is Apple now coming after competing AI assistants on Mac, the only Apple platform they haven’t already kneecapped them on?

Apple knows that most users are lazy, and if they create enough friction, they can make competitors unviable as businesses without outright prohibiting them. That’s exactly what they did in the EU (and were caught) and Epic v Apple (and were found in contempt of court).

We’ll just have to wait and see. I hope it isn’t another evil plan. It would be incredibly foolish for Apple to make such anticompetitive moves right now as DOJ v Apple ramps up towards trial and they stand on increasingly thin ice in Europe.

It’s a disturbingly vague statement that may foreshadow Apple unlawfully kneecapping competing AI assistants to favor its own. We know what genuine Apple permissions look like (e.g. contacts access), and what malicious Apple permissions look like (Apple EU scare screens).

Unclear is this is a genuine push for clearer user permissions requests or a power grab to kneecap competing AI assistants. The two waves of strangely astroturfy posts on the topic don’t bode well, but we’ll have to wait and see.