# Bobby Murphy puts personal money behind Snap's AI video spinoff Dotmo > Source: > Published: 2026-06-18 20:54:16+00:00 [Snap](https://www.snap.com/?ref=runtimewire) co-founder and CTO [Robert "Bobby" Murphy](https://investor.snap.com/governance/leadership-team/default.aspx?ref=runtimewire) will personally lead the investment in Dotmo, a new company being spun out of Snap's internal generative AI video team to build models for interactive gaming experiences, [TechCrunch reported Thursday](https://techcrunch.com/2026/06/18/snap-spins-off-ai-video-team-into-new-company-dotmo-due-to-costs/?ref=runtimewire). The structure is the story. Dotmo is not a clean break from Snap, and it is not a standard corporate venture round. Snap told TechCrunch that Dotmo will be made up of current Snap staff leaving to start the new venture, that Snap will license technology to Dotmo for gaming and interactive entertainment, and that Snap will receive a large equity stake in exchange for the talent and technology license. Snap will not directly fund Dotmo, according to the report, but Murphy will have a significant personal stake while remaining Snap's full-time CTO and continuing to lead Snap's generative AI research and development. Murphy is the right protagonist for this maneuver because Snap's technical bets have long run through him. Snap's own leadership page says Murphy has served as CTO and board member since May 2012 and holds a B.S. in Mathematical and Computational Science from Stanford. Dotmo's named founders, CEO, headquarters, headcount, valuation, product timeline, and outside investors have not been disclosed. The only named backer so far is Murphy, which makes Dotmo less a venture-backed AI startup launch than a founder-led carveout of work Snap no longer wants to finance entirely on its own balance sheet. ### A cost cut dressed as a founder bet Snap gave TechCrunch the plain reason for the move: the high cost of doing this work internally. That explanation tracks with Snap's current financial posture. In Snap's first-quarter 2026 report, Snap reported 483 million daily active users, $1.53 billion in revenue, and $1.60 billion in total costs and expenses for the three months ended March 31. Research and development alone was $478.3 million in the quarter, up from $424.2 million a year earlier, according to Snap's [Form 10-Q](https://www.sec.gov/Archives/edgar/data/1564408/000156440826000027/snap-20260331.htm?ref=runtimewire). Snap also moved aggressively on headcount this spring. On April 15, Snap announced a plan to reduce global headcount by about 16% of full-time employees, with estimated pre-tax charges of $95 million to $130 million, according to Snap's [8-K filing](https://www.sec.gov/Archives/edgar/data/1564408/000119312526155861/d36756d8k.htm?ref=runtimewire). In a separate Q1 investor letter filed with the SEC, Snap said the April action was expected to reduce annualized cost structure by more than $500 million in the second half of 2026. Dotmo fits that same operating discipline, but with a different answer than a layoff. Snap is letting a group of technical employees keep building a high-cost, speculative AI video product outside Snap, while Snap keeps a claim on the upside through equity and a technology license. For the departing staff, Dotmo offers a more natural venture structure for a market where model training, inference, and product iteration can consume capital before revenue is obvious. For Snap, the arrangement preserves optionality without asking public-market investors to absorb the full cost of a non-core AI gaming push. ### Why gaming, not Snapchat Dotmo's stated target is narrower than the generic AI video category. Snap told TechCrunch that Dotmo will focus on AI models capable of creating interactive gaming experiences, and that the licensed Snap technology can be adapted for gaming and interactive entertainment platforms. That distinction matters. Snap has already experimented with generative video for creators. In 2024, Snap introduced an [AI video generation tool for creators](https://techcrunch.com/2024/09/17/snap-is-introducing-an-ai-video-generation-tool-for-creators/?ref=runtimewire), and Snap later announced [Animate It](https://newsroom.snap.com/introducing-animate-it?ref=runtimewire), a Lens+ feature for generating personalized videos from prompts. Dotmo, by contrast, points away from creator clips and toward interactive worlds, a harder product category that overlaps with gaming, simulation, and what some AI labs describe as world models. That is also why the spinoff form makes sense. A Snapchat feature has to earn its place inside a massive consumer app, tied to retention, ads, subscriptions, creator tools, or augmented reality. Dotmo can pursue a different customer and product clock. If the technology becomes useful to Snap later, Dotmo could operate as a partner. If it does not, Snap has separated the experiment from its core operating plan. ### Snap's second carveout in six months Dotmo is Snap's second major spinoff effort of 2026. In January, Snap [spun off Specs into a standalone company](https://techcrunch.com/2026/01/28/snap-gets-serious-about-specs-spins-ar-glasses-into-standalone-company/?ref=runtimewire) focused on Snap's smart glasses line. Two days before the Dotmo report, TechCrunch reported that Snap's new [Specs](https://techcrunch.com/2026/06/16/snap-finally-debuts-its-long-awaited-ar-glasses-specs-and-oof-they-arent-cheap/?ref=runtimewire) would cost $2,195, with preorders opening June 16 and shipments expected in the fall in the U.S., U.K., and France. The parallel is useful but imperfect. Specs is still closely tied to Snap's long-running augmented reality identity. Dotmo is more clearly adjacent: AI models for interactive gaming are not currently part of Snap's core business priorities, a Snap representative told TechCrunch. That makes Dotmo a cleaner example of Snap trying to turn internal R&D into a venture-style option. Public companies have a narrow tolerance for experiments that are expensive, technically uncertain, and far from monetization. AI video is exactly that kind of category. Runway raised a $315 million Series E at a $5.3 billion valuation in February, according to [TechCrunch](https://techcrunch.com/2026/02/10/ai-video-startup-runway-raises-315m-at-5-3b-valuation-eyes-more-capable-world-models/?ref=runtimewire). Higgsfield, founded by former Snap generative AI head Alex Mashrabov, sold another $80 million of stock in a Series A extension and reached a claimed $1.3 billion valuation in January, according to [TechCrunch](https://techcrunch.com/2026/01/15/ai-video-startup-higgsfield-founded-by-ex-snap-exec-lands-1-3b-valuation/?ref=runtimewire). Those rounds show why Snap may prefer to let Dotmo finance itself in the venture market rather than compete internally for budget against Snapchat, ads, subscriptions, and Specs. ### What is still missing The unanswered questions are material. Dotmo has not disclosed a round size, valuation, equity split, compute arrangement, model capability, training data approach, commercial product, or outside backers. Snap's equity stake is described as large, but no percentage has been reported. Murphy's personal investment is described as significant, but no dollar amount has been disclosed. That leaves Dotmo with a clear thesis and an incomplete company profile. Murphy is putting his own capital behind a group of Snap technologists leaving to build outside the walls. Snap is keeping exposure to a category that could matter to games and entertainment, while reducing the cost burden of housing that work internally. The founder logic is unusually direct: when a public company cannot justify carrying an expensive frontier AI team inside the core business, the technical founder can still decide the work is worth backing himself.