Hacking of street cameras is now the de facto modus operandi to obtain intelligence. That includes surveilling ground conditions, traffic flow, criminal activity, and the movement of top leaders. See that in the context of the sea of cameras dotting the globe, across trafficways, roads, alleys, offices, shops, and homes, and you will soon realise how vulnerable advances in artificial intelligence (AI) have made all of us, with the easy sifting of vast video feeds for targeted data.
The evolution in the use of passive cameras installed for deterrence into active networks for real-time surveillance has been dramatic. Russia did this before the Ukraine invasion on February 24, 2022, Hamas hacked Israel’s networks before the deadly attacks of October 7, 2023, and Israel showcased their espionage capability while obliterating the Iranian top deck on February 28 this year. You would be right to therefore assume that the intelligence agencies of other countries and rogue actors with wrong motives know everything about many Indian cities.
Technology adoption: At the heart of this is the use of advanced technology. AI and computer vision algorithms can scan thousands of video streams simultaneously, spotting individuals as well as unusual behaviours. Companies like Palantir and Babel Street apply machine learning to figure patterns, especially suspicious ones. A Palantir app even comes up with geographic deportation zones on a map and dossiers on individuals. Real-time solutions to map facial topology make anonymity in public spaces a myth. Automatic licence plate readers do the same for vehicle tracking, some also capturing other distinguishing features like stickers and dents.
Many of the devices installed are vulnerable as the safeguards are minimal or non-existent — including very basic passwords or none at all, ancient software, versions that haven’t been updated, unpatched security flaws, factory default passwords still in use, and over-reliance on indifferent service providers.
Collateral outcomes: The profound psychological outcome is that once individuals become aware of being constantly watched, the majority prefer to conform, to avoid becoming a part of a database permanently. As Larry Ellison of Oracle said: “Citizens will be on their best behaviour because we’re constantly recording and reporting.”
And when the software misbehaves, or biases emerge from algorithms it’s trained on, it could lead to innocent citizens or groups being targeted. Like the misreading of number plates leads to unwarranted checks and false arrests. This highlights the unintended consequences of proactive policing, the bypassing of democratic oversight, thinning the line between preemptive safety and deliberate population management.
Civil liberty groups are pushing back, demanding the banning of facial recognition in public places, prescribing mandatory data-purging time frames, and wanting warrants for location logs.
With reason. For example, China reportedly has the most number of surveillance cameras, many with AI and facial recognition capabilities, to enforce civil order. But it’s when commercial defaulters are called out on public electronic billboards in authoritarian regimes that it becomes an overreach.
Jonathan Penney explains control, repression, and conformity in the digital age, in his 2025 book, Chilling Effects. He also talks about the misuse of big data by governments and corporations, how it leads to the erosion of democracy and induces compliant behaviours over time.
A February 2026 blog by the American Civil Liberties Union of Massachusetts, a civil rights and liberties organisation founded in 1920, talks about the creation of digital police states and voices concerns about the government surveillance of immigrants, extending to citizens in the phase that follows. Compared with modern AI enhanced tech, the wiretapping, intercepting of letters, and the relentless tracking of informants and alleged communists in the US in the 1950s and ’60s pales significantly.
Coverage span: According to Visual Capitalist, New York, London, Moscow, and Sydney have 10, 13, 20, and 5 cameras per 1,000 people. In comparison, Hyderabad has 79 per 1,000, Indore has 72, Bangalore has 41, Delhi and Chennai have 9, and Pune, Kochi, and Lucknow have 7. While details of individual Chinese cities are not available, estimates are that they have many hundred cameras per 1,000 people, the most in the world.
Repurposing gear for surveillance isn’t uncommon. Like smart streetlights being fitted with acoustic sensors to detect gunshots and Wi-Fi sniffers that track foot traffic. Or the stealth deployment of high-frequency noise machines near cameras to disperse unwanted crowds.
CCTV finds use in various industries for diverse purposes, including turning passive security into an efficiency and profit driver. Like for tracking the gaze of shoppers to reconfigure store layouts that optimise shopping paths in retail, for quality inspection on assembly lines in manufacturing, or for scanning bar codes on conveyor belts to track inventory dynamically in warehouses.
Our Drugs Consultative Committee has mandated the deployment of CCTV at all medical stores, recommending that cameras be placed such that every transaction is verifiable. This was in response to concerns flagged by the National Commission for Protection of Child Rights over the unregulated sale of certain drugs to children.
A call to action: Recently, the Indian government has implemented mandatory STQC (Standardisation Testing and Quality Certification) approval for all internet-connected CCTV cameras, requiring the upfront disclosure of the hardware architecture, origin of chipsets, firmware, software, cybersecurity, vulnerability assessment tests, and supply chain information.
But to avoid dystopian paranoia, more needs to be done. Can we anonymise visual data at the source, so the focus remains on the event rather than the person? Can we mandate a judge-approved warrant, like we need for a physical house search, for querying historical video data? Can we insist on mandatory data-purge timelines? Can our cities maintain registries so we know where municipal surveillance cameras are located?
It’s time to democratise surveillance.