Mo Vij is leaving capability on the table on purpose, and that should make you uncomfortable about your own decisions.

Vij, CEO of 365Labs, is deliberately limiting how AI gets used inside public safety software. Not because the technology cannot do more. Because he has decided it should not, at least not in those contexts, at least not yet. As theadvocate.com reported this week, he is drawing a hard line and holding it, even as competitive pressure to do otherwise stays real and constant.

If you are a founder running AI in your own products or operations, Vij just asked you a question you probably have not answered clearly: are your AI decisions actually yours, or are you just following the current?

The part nobody wants to say out loud

Most founders implementing AI right now are doing it because it feels mandatory. Someone on the team mentioned it, a competitor shipped something, an investor asked about it on a call. So the decision gets made, but not really made. It gets adopted. That is a different thing entirely, and the gap between those two words is where your accountability actually lives.

Vij's move is striking precisely because it runs counter to that pattern. He looked at a category, assessed the risk profile of expanded AI use, decided it did not meet his standard, and chose restriction over capability, governance over feature velocity. That is a founder making a real decision, with real stakes, from a defined position.

When is the last time you did that?

Governance is not a policy document; it is a pattern of decisions

Here is where most founders get this wrong. They think governance means having written rules about AI use, a policy that lives in Notion or gets handed to legal. That is documentation. Governance is something else entirely. Governance is the consistent logic behind your decisions, the thing that makes your choices predictable and defensible over time.

If your AI governance is real, you should be able to answer these without pausing:

  • Which parts of your product or operation are AI-assisted, and which are explicitly not?
  • What standard does a new AI application have to meet before you implement it?
  • Who has the authority to say no, and has that ever actually happened?
  • What would have to be true about a use case for you to deliberately restrict AI access to it?

If any of those made you hesitate, you do not have governance. You have momentum. And momentum is fine until something goes wrong, and then it becomes liability.

The cost of not having a position

Founders without a clear AI position are not neutral. They are reactive. Every decision gets made in the moment, shaped by whoever is in the room, whatever the competitive narrative is that week, or what a vendor just demoed. That is not leadership; it is drift, and it compounds in ways that are genuinely hard to reverse.

Teams that do not see a clear position from leadership start making their own calls. Some of those calls are fine. Some are not. You will not always know which is which until there is a failure, a customer complaint, a regulatory question, or a moment where someone asks you to explain your AI approach and you realize you cannot do it concisely because you have never actually articulated it.

That is not a technology problem. That is a leadership accountability problem wearing a technology costume.

What Vij's decision actually models

Set aside whether you agree with his specific call on public safety. The structure of what he did is worth studying. He identified a high-stakes context. He assessed the gap between what AI can do and what it should do in that context. He made a deliberate call to restrict, and he is holding that position publicly.

That sequence requires three things most founders skip: a defined risk framework, the willingness to say no to capability, and the confidence to defend that decision externally. The third one is where it usually falls apart. It is easy to have principles in private. It is harder when a competitor ships something and a customer asks why you have not.

Vij is holding the line publicly. That is the part that matters most, for what it signals to his team, his customers, and the market about where 365Labs actually stands.

Building your actual position

You do not need a 40-page AI policy. You need a clear, defensible answer to one question: what standard does an AI application have to meet before we use it, and who enforces that?

Start there. Map out where AI currently touches your product, your operations, your customer experience. For each one, ask whether it met a standard or just happened. Then identify the areas where you have not made a call yet, the places where AI could go but you have not decided whether it should.

Those undecided areas are where your accountability gap lives. Not in the things you have already shipped. In the things you have not defined.

A real governance position does not mean being restrictive. Vij is restrictive in one category and presumably expansive in others. The point is not caution for its own sake; the point is intentionality. Knowing why you are doing what you are doing, being able to say it plainly, and having a mechanism that keeps the logic consistent even when you are not in the room.

If you want to get clear on where your AI decisions are actually coming from and build a governance framework that holds under real pressure, that is exactly the work we do with founders inside A&A. Book a session with Mike and let's map it out.

Source: theadvocate.com