【2018 Archives】

When it comes to tech018 Archives Europe has been on a slightly separate track from the U.S. for a while, but the paths are about to diverge big time.

As of today, Aug. 1, the European Union's AI Act is entering into force. The AI Act is a set of rules and regulations designed to protect European citizens from the perils of artificial intelligence.

The full text of the AI Act is available online, but the gist of it is this: it sorts AI applications into several risk categories, with some, such as government-run social scoring, being deemed unacceptable. High-risk AI systems, including those used in critical infrastructure and law enforcement, are regulated, with providers having to follow a strict set of rules on risk management, incident tracking, and more. Providers of AI systems designated as limited risk, including chatbots and deepfakes, are obligated to inform users they're interacting with AI.

In contrast, the Biden administration has issued a set of AI-related safeguards, with numerous, large tech companies on board, but the rules are non-binding, and there's no penalty if they're not followed.

While Europe's AI rules aren't really becoming effective until February 2025, with some of the provisions applying as late as August 2026, the practical outcome for users right now is that the U.S. tech giants are extremely wary of offering AI features to European users.

In June, Apple introduced a set of AI features (dubbed Apple Intelligence), which should go live sometime in the fall, alongside the public launch of iOS 18, iPadOS 18, and macOS Sequoia. And while the AI features will seemingly arrive a little later than expected, some of them have just gone live in Apple's latest batch of developer betas — unless you're in Europe, that is.

In fact, Apple has already officially delayed its AI features in Europe due to regulatory concerns, likely until 2025.

There have been instances in the past of Europe not getting certain features due to the EU's regulators being a tad more stringent on certain matters than those in the U.S.; Meta's social network Threads arrived in Europe months after launching overseas, for example.

But the implications of AI features being delayed in Europe could be far more significant.

In its announcement of Apple Intelligence, Apple said the AI features are "deeply integrated into iOS 18, iPadOS 18, and macOS Sequoia." They're certainly important enough to warrant an entire introductory web page on Apple's website. Apple is not alone in this; if you've seen a tech keynote in the past year or two – any keynote – you've likely heard the word "artificial intelligence" hundreds of times.

• OpenAI rolls out ChatGPT's new Voice AI (without Scarlett Johansson mode)
• This new tool can tell you whether AI has stolen your work
• Instagram now lets you create an AI version of yourself
• Apple's AI features will be late, report claims
• A Google AI model is improving climate forecasting

Individual features, such as Meta's recently launched AI Studio, or ChatGPT's Memory, which are also not available in Europe, are one thing. But Apple touts AI as a key element of the new iPhone, one of the most popular phones in Europe and the world. When the new iPhone 16 arrives in Europe without any of the promised AI features, it's a bigger deal.

It's not just about creating a fun Genmoji image, or Safari summarizing an article for you. Upcoming AI features on the iPhone also include smarter search and editing in Photos, AI-powered spelling and grammar checks, and a significantly smarter Siri, among other key upgrades. A brand new iPhone without any of these (for a couple of months at least) sounds like a very different deal than one with all these features enabled.

Tech giants such as Apple and Meta certainly won't simply give up on AI in Europe. When it announced its AI features will arrive late in Europe, Apple said it will try to find a way to "deliver these features to our EU customers without compromising their safety," and the AI stuff will certainly come in one form or another. But this is just the beginning; as AI becomes more deeply integrated with the devices we use and rely on, it might get increasingly difficult to make them work similarly in the U.S. and Europe, given the different rulesets.

As someone who has lived in Europe most of my life, I'm used to certain features not being available — or being offered in a slightly different form than they are in the U.S. — and I do think that regulating the way big tech uses AI is necessary, and not just in a pinky-swear-not-to-do-bad-things type of way. But I also wonder just how big this rift is going to become now that the EU AI Act is in place.