Samsung Galaxy Watch Blood Pressure: Is It Worth Setting Up?
Apple Quietly Killed Its AI Health Coach: What It Means Now
Apple built a team of physicians, sleep experts, nutritionists, and mental health professionals to train an AI health coach. Then the head of Apple Services looked at Oura and Whoop and decided it wasn’t good enough to ship.
That’s the story of Project Mulberry, and it tells you more about the current state of wearable AI coaching than any product announcement Apple could have made.
In February 2026, Eddy Cue, who took over Apple’s health division, concluded that the planned AI coaching service didn’t meet the bar already set by competitors. The standalone subscription product (sometimes called Health+ internally) is shelved. Features that were in development will roll into the existing Health app individually, on Apple’s timeline, with no commitment to when any specific piece arrives.
This wasn’t a quiet cancellation Apple wanted public. The details came out through Bloomberg’s reporting. The framing: Apple is pivoting, not giving up. The reality: the AI coaching product users were waiting for isn’t coming.
What Mulberry Was Actually Building
Project Mulberry wasn’t a chatbot bolted onto HealthKit.
The internal plan involved AI models trained with clinical expertise across four domains: training coaching, sleep optimization, nutrition guidance, and mental health support. Educational video content was part of the package. Apple was building something that felt less like a data dashboard and more like a health team you could consult on demand.
The subscription tier (Health+) would have been Apple’s answer to the $5.99/month Oura membership and WHOOP’s $199/year subscription. Given Apple’s 2-billion-device installed base, it would have been the largest AI health coaching distribution network ever launched.
None of that shipped. Eddy Cue looked at what Oura and Whoop already do and decided Apple couldn’t credibly enter the market with v1.0.
That decision is worth sitting with. This isn’t Apple concluding the market isn’t valuable. Apple specifically named Oura and Whoop as the quality standard they couldn’t match. That’s a significant public signal from a company that usually says nothing.
The Competitive Implication Nobody Is Saying Clearly
When Apple’s head of services says your competitors set the bar you couldn’t reach, those competitors win on two levels.
First, the obvious one: Oura and Whoop don’t have to compete with Apple Health coach this year or next. That threat is gone. They can keep iterating without an 800-pound gorilla entering their specific niche with a full platform to bundle it into.
Second, the less obvious one: Apple just gave both companies the most powerful third-party validation in consumer technology. You can say “even Apple thinks we’re better” and you’d be accurately representing Eddy Cue’s stated position. That’s not a marketing angle Oura or Whoop manufactured. Apple handed it to them.
Garmin benefits too, though differently. iOS 26.3 rolling out support for actionable smart notifications from third-party wearables (including Garmin and WHOOP) is a DMA-compliance feature, but it arrives at a moment when Apple is actively stepping back from the health coaching space. Apple Watch’s ecosystem advantages are narrowing. The platform is opening. The iOS 26.3 notification forwarding update gave Garmin users a meaningful iPhone experience improvement that didn’t require switching to Apple Watch.
The combined message to iPhone users: you don’t need Apple Watch to have excellent wearable health features, and you won’t have Apple coaching software pulling you back toward the ecosystem either.
Why Oura Won This Round
Cue didn’t say WHOOP was better than Mulberry. He cited both Oura and Whoop as products that already do what Apple was building, and the Oura Ring 4 has had a strong 2026.
The Oura Ring 4 app redesign delivered improved trend analysis that connects sleep quality to daytime energy and workout performance over weeks. That’s exactly the longitudinal insight Mulberry was designed to generate. Oura has years of real user data and iteration behind those insights. Apple was starting from scratch, regardless of how many physicians they hired.
Oura’s form factor matters here too. Sleep tracking only works if you actually sleep with the device on. A ring is more comfortable overnight than a watch for most people, which means Oura’s sleep data is based on better compliance than Apple Watch can realistically achieve. That’s a structural advantage Apple can’t match without building a ring, which they’re not.
At $349 for the hardware plus $5.99/month, Oura isn’t cheap. But the value proposition just got easier to articulate: this is the product Apple decided it couldn’t compete with.
Where WHOOP Stands
WHOOP’s position is different from Oura’s.
WHOOP is a recovery and strain monitoring tool before it’s anything else. The Strain score quantifies physiological stress from workouts. The Recovery score tells you how much capacity you have the next morning. The WHOOP passive MSK strength training review showed how that strain tracking applies to lifters, not just endurance athletes. WHOOP captures training load even when you’re not sweating through traditional cardio metrics.
What WHOOP offers that Apple was trying to build: contextual coaching signals. When your HRV drops three days running and your Recovery hits red, WHOOP’s response isn’t “move your rings.” It’s a clear signal to adjust training intensity. That’s basic physiological coaching based on your actual data, and Apple Watch doesn’t do it well.
The WHOOP 5 vs. WHOOP MG comparison covers what separates the hardware tiers if you’re considering which subscription to start with. At $199/year for the basic tier, you’re paying for continuous monitoring and recovery coaching that Mulberry promised but never delivered.
The subscription-device model is WHOOP’s ongoing tension. You stop paying, you stop getting data. But Apple validated the category. That’s worth something when you’re having the “is this worth $199/year” conversation with yourself.
What the Health+ Rollback Means for Features
Apple said Mulberry features would roll into the Health app individually. That framing is doing a lot of work.
Realistically, here’s how this plays out:
Sleep analysis improvements. Most likely to ship first and soonest. Apple Watch already captures the raw sleep stage data. Enhanced analysis and something resembling a readiness score could arrive in iOS 27 or even a 26.x update. Don’t expect it before WWDC 2026.
Training load guidance: less clear. This requires Apple to make recommendations based on workout history and recovery signals, which gets into coaching territory they’ve been cautious about.
Nutrition coaching. The least likely near-term addition. Nutrition recommendations at the app level involve liability considerations, FDA implications given the evolving rules on fitness wearables, and the fact that Apple doesn’t have a food database or logging UI.
Mental health support: probably years out. Apple has made incremental moves here with mindfulness and cycle tracking, but a clinical-quality mental wellness coaching layer requires regulatory navigation that moves slowly.
The practical translation: Apple Watch users will get iterative improvements. Nothing that resembles the unified AI health coaching layer Mulberry was designing. If you’re holding out for Apple to fix this through OS updates, you’re looking at a multi-year wait for features that exist in third-party apps today.
The New Entrants Aren’t Standing Still
The new Fitbit 2026 AI health coach release added another player to the AI coaching space Apple vacated. Google’s health ambitions through Fitbit have their own momentum, separate from what Apple does or doesn’t ship.
WatchOS 26 added Workout Buddy, an on-wrist AI coaching feature narrowly scoped to workout guidance during sessions, not the longitudinal coaching layer Mulberry was targeting. Workout Buddy is a useful addition. It’s not a replacement for what Mulberry would have been.
The AI health coaching space is crowded in 2026 in a way it wasn’t in 2023 when Apple started Mulberry. The delay cost Apple the window when they might have set the standard rather than chased it.
Who Should Pay Attention to This
Current Oura users: You’re in the best competitive position you’ve been in years. Apple’s own leadership cited your product as the benchmark it couldn’t match. The competitive threat that might have pulled your subscribers toward a first-party Apple solution is gone, at least for this product cycle.
Current WHOOP users: Same validation applies. And with iOS 26.3 improving WHOOP’s notification experience on iPhone, the friction of using WHOOP as an iPhone user just dropped. The subscription is easier to justify when the platform you’re using it alongside isn’t actively trying to replace it.
Apple Watch owners waiting for the Health app to get smarter: Adjust expectations. You’ll get incremental improvements on Apple’s schedule. If you’re training seriously and making decisions based on recovery data, a supplemental tracker like Oura or WHOOP fills the gap that Apple acknowledged it can’t close soon.
People considering their first wearable purchase: The field simplified slightly. You’re not comparing options that include an imminent Apple AI coaching subscription. The best AI fitness coach apps for 2026 breaks down the field as it stands now. No ghost product to wait for.
Garmin users who want Apple Watch features: The calculus keeps tilting toward Garmin for serious fitness tracking on iPhone. iOS 26.3 notifications, Garmin’s own coaching depth covered in the 2026 update roundup, and now Apple stepping back from health coaching: the reasons to switch to Apple Watch for fitness reasons keep shrinking.
The Honest Bottom Line
Apple had a credible shot at becoming the dominant AI health coaching platform. The installed base, the health hardware, the brand trust in health contexts: all of it pointed toward an inevitable Apple entry that would reshape the market.
It didn’t happen because the product wasn’t good enough. By Apple’s own assessment.
That’s the sentence that matters. Not that Apple failed. Not that wearable AI coaching is too hard. Apple concluded it wasn’t at the level of products that already exist.
If you’re trying to decide whether Oura, WHOOP, Garmin coaching features, or one of the AI coaching apps is worth paying for: the answer is yes, and Apple just inadvertently told you so. The alternatives aren’t second best while you wait for Apple. They’re the current standard that Apple couldn’t match.
Pick the one that fits how you train. Oura Ring 4 if sleep and recovery data is your priority. WHOOP if you push hard and need strain quantification. Garmin if you’re a runner or cyclist who wants on-wrist coaching and weeks of battery life. The right choice depends on your training, not on waiting for Apple to ship something.
Reporting based on Bloomberg’s coverage of Project Mulberry and Apple’s February 2026 health division changes, cross-referenced with Apple’s iOS 26.3 release notes for third-party wearable support. Prices current as of March 2026.
