Snap’s AI video team is leaving to start their own thing — and it’s probably about money

Snap is doing what big tech companies do when they realize a promising internal project is getting too expensive to keep under one roof: they’re spinning it off.

The team working on AI video inside Snapchat’s parent company is packing up and forming a new outfit called Dotmo. The staff involved are leaving Snap proper to focus entirely on building AI-powered video tools. This isn’t a layoff — it’s a separation, and a calculated one at that.

Snap has been through this before. They spun off a drone division a couple years back, and that didn’t exactly set the world on fire. But AI video is a different beast. The space is red hot, and the capital flowing into generative video startups right now is absurd. Runway, Pika Labs, and a dozen others are burning through venture money like it’s going out of style. Snap knows they can’t compete with that spend level internally, especially when they’re still trying to make Snapchat’s ad business work.

So Dotmo gets to go out into the wild, raise its own funding, and build whatever it wants without Snap’s quarterly earnings calls looming over every decision. The parent company likely retains some stake or licensing rights, but the details there are murky. What’s clear is that Snap gets to trim costs and still keep a finger in the AI video pie.

The timing makes sense. Snap’s stock has been on a rollercoaster for years, and the market is punishing any company that can’t show a clear path to profitability. Keeping a speculative AI video R&D team on the payroll while trying to convince investors you’re serious about margins is a hard sell. Letting them go form their own company is the kind of move that keeps Wall Street from asking too many questions.

For the team, it’s probably a relief. Internal projects inside big social media companies have a way of getting bogged down by politics, shifting priorities, and the constant question of “how does this help us sell ads?” Dotmo gets to skip all of that and go straight to building. And if they build something impressive, the exit options are plenty — acquisition by a larger AI player, an IPO down the line, or just staying independent and profitable.

The AI video space is still wide open. Text-to-video generation is getting better by the month, but nobody has truly cracked real-time, high-quality, controllable video at scale. If Dotmo can pull that off, they won’t have trouble finding customers or investors. Snap’s loss could be their gain.

I’m curious to see what Dotmo actually ships. A lot of these spin-offs end up being quiet exits or get absorbed back into the parent company after a few years. But the AI video market has real momentum, and the talent leaving Snap has been working on this stuff internally for a while. They’re not starting from zero.

This is a smart move for everyone involved. Snap gets to focus on its core business, the team gets freedom and funding potential, and the rest of us get another player in the AI video arms race. I just hope Dotmo’s first product isn’t another “generate a 3-second clip of a cat wearing a hat” demo. We’ve seen enough of those.