Why a French surveillance startup is targeting Britain to deploy AI tech banned at home
A French artificial intelligence company is quietly expanding across Europe with technology its domestic market won't allow. Vizzia, which has just raised €30m in a Series B round led by Silicon Valley's Base10 Partners, has built AI-powered surveillance cameras for municipalities — but France's ban on real-time algorithmic detection means the startup is looking elsewhere to deploy its full capabilities.
The target markets? Britain and Italy, where regulations are, in the company's own words, 'different'. What cofounder Katrin de Proyart means is looser. While Vizzia's cameras in France must rely on human operators to review footage after the fact, the company is positioning itself to offer more automated surveillance in jurisdictions that haven't imposed similar restrictions. Total funding now stands at €50m.
This is regulatory arbitrage in action. A startup hamstrung by privacy protections at home is jurisdiction-shopping across Europe, seeking municipalities eager for surveillance tools without the legal guardrails. The involvement of two American venture capital firms — Base10 and Headline — alongside French investor Sistafund suggests international investors see opportunity in Europe's fragmented approach to AI governance.
The French paradox
France currently prohibits real-time AI-powered detection in public spaces, a stance that sets it apart from many European neighbours. Vizzia's technology can flag suspicious movements or behaviours, but domestic regulations require human review before any action is taken. The cameras capture footage; algorithms can analyse it, but only retrospectively.
Yet even France's position is softening. Legislation passed in 2023 temporarily authorised real-time AI surveillance during the 2024 Paris Olympics. That exemption has since been extended through 2027, a sign that political appetite for these tools is outpacing privacy concerns. What was framed as a temporary measure for a major event is becoming semi-permanent policy.
This creates an awkward situation for Vizzia. The company initially focused on detecting illegal waste dumping, deploying cameras to fly-tipping hotspots across 250 French municipalities. Last year it expanded into broader crime and antisocial behaviour monitoring. But the full potential of its AI systems — real-time alerts that could trigger immediate intervention — remains locked behind French law.
So the company is going where it can operate with fewer constraints. Britain and Italy offer what de Proyart describes as markets 'where regulations are different', allowing Vizzia to tailor what its systems detect based on local rules. The implication is clear: where laws permit more invasive surveillance, Vizzia will provide it.
What the polls don't tell you
De Proyart justifies the expansion by citing internal research. 'We've done many studies on the relevance of video surveillance and our polls are clear: almost all political parties want access to these tools,' she says. The framing is instructive — demand presented as politically universal, safety positioned as the overwhelming priority.
But these are Vizzia's own polls, conducted by a company with a commercial interest in demonstrating demand for its product. The claim that 'almost all political parties' support expanded surveillance deserves scrutiny. Which parties? Across which jurisdictions? What questions were asked, and how were they framed?
What's interesting here is how surveillance adoption is being sold as purely responsive to public need, rather than driven by vendors seeking markets for profitable technology. De Proyart describes 'incredible momentum' in the sector, with safety 'rapidly becoming a priority' for local authorities. Perhaps. But the presence of well-funded startups actively marketing these tools to cash-strapped councils might also be accelerating that momentum.
The broader European picture complicates this narrative. The EU's AI Act, currently rolling out in stages across member states, explicitly bans using CCTV footage to create facial recognition databases. But enforcement remains patchy, and compliance varies significantly between countries. Vizzia insists its products are 'compliant by design', built to conform with existing regulations. Digital rights organisations are less convinced that current laws are adequate to protect citizens from mass surveillance creep.
An expanding market
Vizzia isn't alone in spotting opportunity in Europe's patchwork regulatory environment. Paris-based Orasio raised a €16m seed round last year for AI software that analyses video feeds to detect weapons or fires. The sector is attracting significant venture capital despite — or perhaps because of — regulatory uncertainty.
American investors appear particularly enthusiastic. Base10 Partners and Headline bring Silicon Valley capital and expectations to a market where privacy protections remain stronger than in the United States, but where enforcement is fragmented and political will is wavering.
With its fresh funding, Vizzia plans to open offices in Britain and Italy, targeting one new local authority client per week throughout 2026. The company will expand from 100 employees to 250 by year's end, a significant hiring spree focused on commercialisation and research and development.
That R&D investment raises questions about where the technology is heading. If Vizzia is building more sophisticated detection capabilities now, whilst regulations remain uncertain, what happens when — or if — those tools become legal in more jurisdictions?
France's temporary Olympic surveillance measures are being extended for three more years. Britain's regulatory approach remains unclear as the government considers its position on AI governance. Italy is implementing EU directives at its own pace. The result is a continent where the rules governing public space surveillance vary dramatically depending on which side of a border you're standing on.
For Vizzia, that fragmentation represents commercial opportunity. For European citizens, it means the extent of algorithmic monitoring in their daily lives depends less on democratic consensus than on which regulatory environment their local council happens to operate within. As venture-backed surveillance startups scale across jurisdictions, that inconsistency looks less like federalism and more like a loophole.
