The personal AI assistant race has a new runner, and it does not want you to open an app. On October 7, a startup called Tab emerged from stealth at a $300 million valuation. Its pitch is simple: you don’t visit Tab. You text it, on iMessage or WhatsApp, and it does the thing.
Ordering groceries. Finding a birthday present. Booking travel. Making the phone call you have been putting off for three weeks. You text Tab the way you would text someone you trust, and it takes the weight off. That is the product, the whole product, and it might be the most interesting interface decision in the current AI race.
Deliberately boring mechanics
There is no new app to learn and no new habit to form. You message Tab inside the threads you already use every day. Behind the messages, Tab has its own phone number, its own computer, and its own wallet. It does not tell you how to do the thing. It does it.
Two capabilities carry the demo. For phone work, Tab can call a hotel, a doctor’s office, a restaurant, or a vendor, wait on hold, ask the questions, and report back in Messages. For money, the rule is explicit: before anything is spent, you see exactly what it is, the total, and who gets paid. Nothing is paid until you approve it. That is the line between an assistant and a liability, and Tab draws it in the right place.
The company was founded earlier this year by Brennan Erbz, Stafford Schlitt, and Ammar Amdani. Backers include SV Angel, Valar Ventures, and American Spirit. The funding amount was not disclosed, but a $300 million valuation on a just-out-of-stealth consumer AI company tells you everything about how hot this category is. Tab is in private beta, growing mostly by word of mouth, with a new product launch promised soon.
Leading with the hard question
Here is where Tab gets interesting. An assistant that can read your messages, call businesses on your behalf, and spend your money is “only as good as your trust in it.” Those are co-founder Ammar Amdani’s words, and they are unusually honest for a launch. Most AI startups lead with capability. Tab leads with whether you should let it in at all.
The stated safeguards: your data is never used to train AI, your card sits in a vault that Tab itself cannot see, and nothing that cannot be undone happens without your approval. Fair enough as a starting position. The honest caveat is that these are the company’s own claims, relayed in interviews, with no independent testing yet. Trust is a feature you have to ship repeatedly, not announce once.
Still, the framing explains why Tab might have a lane in a race that already includes heavyweights, from breakout consumer assistants to Meta’s own personal agent push. Everybody is racing to make AI more capable. Amdani’s bet is that the winner will not be decided on capability. It will be decided on trust. And the interface people already trust most is the text thread.
Why the bet is smart
Think about the last AI product that asked you to change a habit. Most of them are sitting in your downloads folder, opened twice. The text message is the opposite: the most-used interface on the planet, the one your grandmother and your teenager both check every hour. An assistant that lives there inherits all of that attention without asking for any of it.
Texting someone to handle your errands is how the wealthy have always lived, by delegating to people. Tab is selling the feeling of having a person on it, minus the person. The wait-on-hold detail is the tell: nobody wants a feature that dials the airline. Everybody wants the airline dealt with. A product that absorbs the unpleasant parts of modern life has a demand curve that does not need much explaining.
Whether it earns the valuation comes down to one thing: the first hundred thousand users, and whether they trust it with their wallets. The approval loop is the product’s load-bearing wall. If approvals are smooth and rare, Tab feels like magic. If every grocery order becomes a confirmation gauntlet, it feels like a bureaucrat. The beta is where that balance gets tuned.
Be First to Comment