In 2025, the financial sector continues its courtship with generative AI, hoping it will finally make compliance less of a bureaucratic slog. According to Global Relay’s State of AI in Surveillance Report 2025, attitudes are thawing: there’s been a 19% drop in firms reluctant to implement AI. Apparently, nothing eases doubts like the promise of automated paperwork and plausible deniability. Still, not all is rosy. Explainability remains elusive, regulators are breathing down necks, and integration often resembles a Frankensteinian patchwork. Innovation, it seems, comes with a compliance hangover.

Beyond finance, AI is now the darling of physical security—spotting odd behaviour, identifying faces in crowds, and predicting crimes before they happen (or not). German defence start-up Helsing, for example, has plans to release a swarm of underwater drones with “advanced AI” to enhance naval surveillance. Because nothing says peace of mind like the quiet hum of a robot lurking in the deep.

Convergence of physical and digital monitoring

Surveillance has gone from watching CCTV tapes in a dingy room to something called “visual intelligence,” where AI analyses footage in real-time for anything resembling trouble—or, occasionally, a dog mistaken for a burglar. This AI-augmented vision is increasingly tied into the grand convergence of IT, OT, and security systems—because if you’re going to be vulnerable, you might as well be efficiently vulnerable.

The UK’s Royal Navy is piloting Project Cabot, which aims to have uncrewed undersea vehicles communicate with AI-powered acoustic systems. A noble endeavour, if the goal is to create a sentient navy that pings back.

AI surveillance in 2025: A global panorama

Surveillance in 2025 is more pervasive, more automated, and far more likely to call you suspicious for looking at your phone the wrong way. States and corporations alike are racing to squeeze every drop of insight from your data, sometimes in the name of safety, other times for more creative interpretations of “national interest.” The boundary between defence and intrusion continues to blur, and the average citizen finds themselves in the awkward position of being simultaneously protected and targeted.

Note: The following is hardly a complete list. Global surveillance programmes are as widespread as they are hush-hush, and many remain hidden under layers of classification or charming euphemisms like “public safety initiative.” This curated selection offers a glimpse into the more transparent corners of the global surveillance theatre.

Europe

United Kingdom

In true dystopian fashion, the UK’s Ministry of Justice uses AI to profile over 1,300 individuals daily to determine their risk of reoffending. The next phase? A system to predict potential murderers. What could possibly go wrong?

Belgium

A recent report has dared to suggest that predictive policing might not be entirely ethical—particularly when it relies on dodgy databases and sociological guesswork. Brussels, ever the voice of reason, is considering a ban.

Hungary

Not content with just banning Pride marches, Hungary has weaponised facial recognition to identify and penalise participants. It’s privacy invasion meets political repression, with a dash of algorithmic bias.

Middle East

Saudi Arabia

NEOM, the futuristic $8.8 trillion megacity, promises jetpacks, robot servants, and—of course—total surveillance. Reports of forced displacements and up to 21,000 worker deaths have somewhat dulled the shine, though the PR brochures remain top notch.

Iran

Iranian authorities have turned AI into an instrument of suppression, employing facial recognition and geolocation to crack down on protest. Following Mahsa Amini’s death, over 20,000 were arrested, proving once again that authoritarianism scales well with machine learning.

Asia

China

China remains the gold standard of techno-authoritarianism. AI surveillance is omnipresent, especially in Xinjiang, and its export to other regimes makes Orwell look quaint. Exporting digital repression: now available on Belt and Road.

India

India’s surveillance sector is booming. AI tools are deployed with gusto—even as spyware scandals, Supreme Court probes, and human rights concerns pile up like unread privacy policies.

Africa

South Africa

Gauteng province has partnered with Vumacam to install over 6,000 cameras loaded with AI capabilities like licence plate recognition and behaviour detection. Naturally, there’s no real regulation for this sort of thing—but rest assured, someone somewhere is watching.

Uganda

During the 2021 elections, Uganda demonstrated the dark art of algorithmic suppression by using AI to monitor social media and stifle dissent. Proof that surveillance tech, once acquired, rarely sits idle.

South America

Argentina

Argentina has launched an AI crime prediction unit, apparently inspired by Minority Report and a generous interpretation of due process. It uses historical data to divine future misdeeds—bias included at no extra charge.

Brazil

Brazil’s surveillance infrastructure is powered by algorithms that trawl state databases with minimal oversight. A dream for statisticians, a nightmare for civil liberties.

Oceania

New Zealand

NZ police teamed up with SaferCities to roll out the vGRID platform: a network of thousands of cameras accessible from officers’ phones. During COVID-19, police reportedly flagged cars as “stolen” to bypass lockdown rules. Creative, certainly. Legal? Not so much.

Privacy vs. security tensions

Federated learning and privacy-preserving AI are all the rage in marketing decks. Apple champions on-device processing, which is lovely—until you realise your phone still knows more about you than your best mate.

The US, meanwhile, warns of AI-enabled disinformation, though it’s unclear if the concern is ethical or competitive.

And then there’s Meta’s AI-powered therapy bots. What could be more reassuring than baring your soul to a chatbot owned by the world’s largest advertising firm?

Global power shifts

Surveillance is no longer just about keeping an eye on things—it’s about who controls the eye. The battlefield now spans land, sea, space, and that murky zone where your toaster knows more than you do. And the AI arms race shows no sign of slowing down—especially with authoritarian states writing the rulebook as they go.

Resources