The New York Times is testing an AI feature in its Cooking app that lets subscribers change recipes to fit what they’ve actually got in the kitchen. The Verge AI reports that the launch has already split the app’s loyal user base. NYT Cooking senior editor Genevieve Ko announced the tool last Friday. She described it as a way to “generally make our recipes work for what a cook needs at that moment.”
For now it’s live for all Cooking subscribers on iOS. The Times says wider availability is coming “soon.”
What the AI Recipe Tool Actually Does
Here are the core capabilities, as detailed in The Verge AI:
- Ingredient substitutions that change the method. This isn’t a simple find-and-replace. In one example, a user asks to use baby spinach instead of kale. The AI notes that baby spinach wilts faster, so it tells the cook to add it at a later step. It adjusts the method along with the shopping list.
- Equipment swaps. Don’t have the Dutch oven the recipe calls for? Ko says the tool can rework a recipe around the pans and tools you do have.
- Scaling to any serving size. An older feature could already halve or double recipes. The new one scales to any number of servings, which is handy when you’re cooking for three or feeding a crowd of eleven.
- Private customizations. Your modified recipes stay private. You can’t publish them or share them through the app, so the original editorial recipes stay as the official versions.
- A gradual rollout. The Times is releasing the feature in stages and watching how it performs before it expands further.
AI Isn’t New to NYT Cooking
The Times has used AI in this app before, just with less fanfare. According to Ko, AI converted imperial volume measurements into metric weights across recipes two years ago. Last year, the app added the halve-or-double conversion.
The difference now is visibility. Those earlier uses happened in the background and produced fixed results that editors could check. This new tool is conversational and generates guidance on demand. That’s far harder to quality-control when the publication’s brand depends on tested recipes.
The Backlash
Not everyone wants it. Readers filled the comments on Ko’s announcement with concerns about AI’s environmental footprint and about what AI features do to the app experience overall.
One reader asked if they could turn the feature off. Ko answered: “No, not at this time.” She added that the feature is “completely optional to use.”
Another user was blunter: “Unbelievably disappointing to see. I’ve always loved NYT Cooking and it pains me that the AI mandate has reached what used to be a wonderful and simple collection of recipes and notes from all sorts of different people.”
The comments on the post have since closed. NYT spokesperson Jordan Cohen told The Verge that this is standard practice. He said comments stayed open for four days while Cooking editors answered questions, and that the Times closes comments on older articles so moderators can focus on new discussions. Cohen also said many Cooking users are “already trying it out on our iOS app; they’re reporting good results.”
Why This Matters
What stands out here is how the tension plays out. NYT Cooking built its reputation on human-tested recipes and a busy community of reader notes, where cooks swap tips on exactly these kinds of substitutions. An AI that answers “can I use spinach?” instantly is useful. It also competes with the human layer that made the app worth paying for.
The “optional but not removable” approach is becoming the default across consumer apps. Companies put AI in front of everyone and let usage decide what happens next. That works when the feature stays out of the way. It causes friction when users feel the product’s character has changed.
There’s also a trust question. A tested recipe carries the Times’ editorial guarantee. An AI-modified version doesn’t, at least not in the same way. If a substitution ruins someone’s dinner, it’s unclear who owns the mistake. Keeping custom recipes private looks like a deliberate move to keep AI output separate from the edited catalog.
What Comes Next
Watch two things: how fast the feature reaches Android and web, and whether the Times eventually adds an off switch. Expect other recipe platforms and food publishers to follow if Cooking’s engagement numbers look good, since recipe adaptation is one of the clearest everyday uses for generative AI. The full story and more reader reactions are in The Verge AI’s original report.