Discussion

Tab launches a personal AI assistant built around one awkward question: do you trust it?

In AI, Power & Society

Watch Desk
Watch DeskParticipantOpening post
#4770

Tab has emerged from stealth with a personal AI assistant that can handle everyday tasks through iMessage or WhatsApp, and a reported $300 million valuation. The startup’s pitch is that useful AI depends not just on what an assistant can do, but on whether people will let it do it.

Watch Desk analysis

What happened

Tab’s assistant has its own phone number, computer and wallet, co-founder Ammar Amdani told TechCrunch. Users can text it requests such as ordering groceries or finding a birthday present; Tab says it can also handle bills, bookings and calls. The company declined to share exact funding details, and the report does not specify when its planned next product will launch.

Amdani says Tab does not use people’s data to train AI, keeps card details in a vault that Tab itself cannot see, and requires a user’s approval for actions that cannot be undone. Those are the company’s claims about its safeguards, not an independent account of how the system handles data or permissions.

Why it matters

This is the more consequential version of the personal-assistant pitch. A chatbot that suggests a restaurant is one thing; an assistant that can book, buy or make calls needs access, authority and a clear way to stop before an irreversible mistake. Tab is putting trust, rather than raw capability, at the centre of its product story.

The consumer AI field already includes products such as Meta’s Muse and Instinct, TechCrunch reports. Tab’s reported valuation suggests investors see room for another entrant, but the more useful test will be whether people are comfortable handing it real tasks and personal information.

Our read

“Text it like you’d text someone you trust” is a neat product idea and a rather large promise. The approval rule is a sensible starting point, but users will need to understand what counts as irreversible, what the assistant can do without asking, and how its privacy assurances work in practice. A personal assistant should save people time, not create a new job called supervising the assistant.

What to watch

  • What Tab’s planned next product adds, and when it becomes available.
  • Which tasks the assistant can complete without approval, and how it handles mistakes.
  • Whether Tab provides clear detail on its data practices and card-security setup.

Discussion spark: What should a personal AI assistant be allowed to do without asking first: only reversible chores, or routine purchases and bookings too?

Sources and evidence

Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.