Snap spins off AI video team into new company, Dotmo, due to costs

The Spin-Off Signal: Snap’s AI Video Unit Becomes Dotmo

Snap Inc. is spinning off its AI video development team into a standalone company called Dotmo, marking the second time in recent months that the Snapchat parent has carved out an internal unit. The move, reported by TechCrunch, sees current Snap staff leave the social media company to focus exclusively on AI video technology under a new corporate structure. The stated driver: cost containment.

This is not a mere reorganization. Snap is effectively acknowledging that certain AI research and product verticals—particularly those requiring heavy compute and specialized talent—are better funded and managed outside the constraints of a social media balance sheet. By spinning out the team, Snap offloads the operational burn while retaining strategic ties through equity or licensing agreements. It is a pattern we have seen before: Meta’s spin-off of AI research into FAIR (though internal) and Google’s creation of DeepMind as a semi-autonomous entity.

Why This Matters

The decision underscores a growing tension in the AI industry: the gap between experimental AI video generation and profitable product integration. Snap has been aggressive with AI features—My AI chatbot, AR lenses, and now video generation—but the cost of training and inference for video models is orders of magnitude higher than text or image models. For a company that still relies heavily on advertising revenue and has not yet proven a return on its AI investments, keeping a dedicated video team in-house becomes a liability.

Dotmo’s creation signals that even well-capitalized social platforms are struggling to justify the R&D spend on frontier AI video internally. This could accelerate a trend where specialized AI teams break away from parent companies to seek venture funding, licensing deals, or enterprise customers. For the broader industry, it suggests that the “AI arms race” is fragmenting: large language models may stay in-house, but video generation is increasingly becoming a standalone business category.

Implications for AI Practitioners

For AI engineers and researchers, the Dotmo spin-off offers a clear career signal. The market is rewarding specialization in high-cost, high-compute domains like video generation, but only when paired with a focused business model. Practitioners should expect more spin-outs from large tech firms, creating new startup opportunities—but also greater volatility. Working on AI video at a social media company may now come with the risk of being “spun off” into a separate entity with different equity, culture, and funding dynamics.

Additionally, the move highlights that cost efficiency is becoming a primary driver of AI organizational design. Practitioners should prioritize understanding inference optimization, model distillation, and compute budgeting—skills that will be valued as companies seek to make video AI economically viable.

Key Takeaways

• Snap is spinning off its AI video team into Dotmo, citing cost pressures, reflecting the high operational expense of video AI development.
• This trend may accelerate as other tech firms separate capital-intensive AI research from core social media or advertising businesses.
• For AI practitioners, specialization in video generation offers opportunity but comes with increased organizational instability and a need for cost-aware engineering skills.
• The spin-off signals that the market is moving toward standalone AI video companies rather than keeping such teams inside larger platforms.