By Sandeep Budki
India’s rapid expansion of surveillance camera networks under smart city, transport and public safety programmes is sharpening scrutiny over how AI is deployed, secured and governed, as authorities look to move beyond basic video recording toward real-time monitoring and response. Urban local bodies, police departments and transport agencies have rolled out large numbers of cameras across roads, transit hubs and public facilities. However, much of this infrastructure still operates in a record-and-review mode, where footage is analysed after an incident rather than informing live decision-making.
“India is at a point where AI-enabled video intelligence is both a need of the hour and a preparedness imperative for the future,” said Nakul Duggal, EVP and Group GM for automotive, industrial and embedded IoT and robotics at Qualcomm Technologies. While the country has one of the fastest-growing CCTV markets globally, real-time intelligence remains limited. The challenge for the operator is the lack of real-time understanding of the video they generate. However, edge-based processing changes how surveillance functions; AI at the edge ensures that video is processed in real time on the device or an on-premises server.
With the rapid shifts happening across sectors from dense urban environments to proactive public infrastructure and enterprise needs, AI-enabled surveillance has become essential, not optional. Duggal said that this will form the base for the next decade of video intelligence.
Homegrown security and surveillance brand CP Plus echoes this assessment too, positioning AI-based surveillance as critical to the present security and governance needs. Aditya Khemka, MD, CP Plus, said, “AI in surveillance is a crucial necessity for India today, not a speculative future trend or a discretionary upgrade,” pointing to the limitations of traditional systems as cities grow and infrastructure becomes more complex. AI changes the role of cameras from passive recorders to active systems capable of perception and contextual understanding, allowing authorities to intervene earlier rather than respond after an incident. This shift is especially relevant for traffic management, perimeter protection and dispute resolution, where speed and accuracy influence outcomes.
Both companies emphasise on-device AI as central to securing camera-based recognition systems. “On-device AI is central to securing the next generation of camera-based recognition systems,” Duggal said, adding that authentication and decision-making at the edge make systems more resilient. Edge processing helps identify spoofing attempts before data leaves the device and ensures that “sensitive video does not travel across networks for analysis.
According to Khemka, on-device AI is crucial for maintaining the authenticity and security of camera-based recognition systems. While local processing limits external exposure, “by running AI models directly in cameras or edge hardware, recognition processes are shielded from external networks, reducing the risk of cyber manipulation, signal interception, or synthetic spoofing.”
As AI-led surveillance expands, questions about the standards defining acceptable and unacceptable uses of camera AI in public or semi-public settings are also gaining prominence. Duggal said that clarity around usage is needed to build trust. “Our job is to assist them by providing platforms that are built for operational sovereignty, privacy-aware processing, and secure on-device intelligence.”
Khemka said India would benefit from clearer standards. “Current legal frameworks provide general data protection provisions but do not sufficiently address how visual AI systems should be governed in real-world environments. The introduction of well-defined standards for the use of camera AI in public and semi-public spaces will help in the maturation of India’s surveillance ecosystem, and such guidelines would help align authorities, technology providers and citizens.”
