FiveAI raises £14M Series A and £12.8M as part of UK government's StreetWise Project for self-driving tech, is using CCTV data to tackle prediction
Matt Burgess / WIRED UK :
Context & Ripple Effects
FiveAI is stacking two very different balance sheets at once: a £14M Series A from private investors alongside £12.8M from the UK government's StreetWise Project, making the state a direct funder of its self-driving software rather than just a regulator. The CCTV-data angle is the technical differentiator — FiveAI is mining existing urban camera networks to build the prediction models that let a car anticipate pedestrian behavior, instead of relying solely on onboard sensors.
This puts FiveAI at the front of a small UK cluster. The same StreetWise project later graduates into public-road trials on a 19km London route with safety drivers, and FiveAI goes on to raise a $41M Series B by 2020 — evidence this dual-funding start held. The backdrop is a funding asymmetry: as Reuters' survey of Oxbotica, FiveAI, and Wayve frames it, these startups are targeting EU-specific driving challenges to compete against far better-capitalized US rivals.
First-order effects
- FiveAI enters 2017 with roughly £26.8M across private and public sources, enough to fund its CCTV-based prediction research while peers must choose between investor money or grant money.
- The UK government becomes an active participant in self-driving development through StreetWise, not merely a rules-setter — its money now rides on FiveAI's technical bet paying off.
Second-order effects
- Rival UK autonomy firms Oxbotica and Wayve now compete against a partially state-funded peer, sharpening the case for their own differentiated approaches — Wayve's learning-based system eventually draws a $1.05B SoftBank-led Series C, the UK's largest AI fundraise.
- Using CCTV footage for training data pulls surveillance infrastructure into commercial AI development, forcing questions about consent and data governance that sensor-only competitors don't face.
Third-order effects
- If the pattern holds, national governments become standard co-investors in autonomy stacks — capital policy doubling as industrial policy for who owns the self-driving layer in each region.
- Repurposed public camera networks as training corpora could normalize city-owned data as an AI asset class, with governance frameworks lagging behind the deals that exploit it.
The trend: UK autonomous-driving startups are closing a funding gap with US rivals by pairing venture rounds with government program money, turning state involvement into a competitive moat.