Training an AI diagnostic model can take weeks. Getting regulators to approve it can take 12 to 18 months, and that mismatch is now a bottleneck for the entire medical AI sector.
Scarlet, a London-based certification body for AI and software medical devices, has raised a $50M Series B to attack the gap. Headline led the round, joined by new investor Notion Capital and existing backers Atomico, Creandum, Kindred Capital and Entrepreneurs First. Headline partner Nancy Xiao Mclaughlin joins the board.
The company positions itself as a specialist reviewer sitting between device developers and national regulators, handling the technical assessment work that generalist notified bodies were never built to do quickly.
The funding supports expansion into the United States and Asia-Pacific, the two markets where AI device submissions have grown fastest. Both regions have their own approval pathways, which means the company needs local expertise rather than a single global process.
The commercial logic is straightforward. Every month a validated algorithm spends waiting costs its developer revenue and gives competitors room to move. Certification speed has become a competitive factor in its own right.
Scarlet’s raise lands as European and US regulators race to write frameworks for continuously learning devices, a category where traditional one-time approval may not be workable.
