Artificial intelligence is already an integral part of today's security and audiovisual industries.
Far from being an emerging technology, its integration into video surveillance, access control and video analytics systems is transforming the market and creating new opportunities for systems integrators, installers and managed service providers.
According to market data, the AI-powered video surveillance market reached a value of USD 6.51 billion in 2024 and is expected to exceed USD 28.76 billion by 2030, growing at a compound annual growth rate (CAGR) of 30.6%. According to Anna Acopian, Founder and CEO of SafegateAI and moderator of the Integrate AI Conference, this reflects a market reality that is already shaping investment decisions across the industry.
Beyond security: AI's role in energy efficiency
One of the most promising—and still underexploited—applications of artificial intelligence is its ability to improve energy management in buildings.
Permanent video surveillance infrastructures continuously generate vast amounts of data. By adding AI-powered analytics, that same infrastructure can become a platform for optimising a building's energy consumption.
A study published in Energy Informatics (Springer, 2025), based on an analysis of 126 research papers, found that reinforcement learning systems can reduce energy consumption in smart buildings by 22.3%, while hybrid AI solutions achieve savings of up to 28.1%. The return on investment ranges from 2.1 to 5.8 years.
In practical terms, a surveillance camera installed in an office building can detect that an entire floor is unoccupied during the night and automatically notify the building management system to reduce HVAC operation. The infrastructure is already in place; the added value comes from the artificial intelligence layer.
Since buildings account for approximately 36% of global energy consumption and nearly 40% of global CO₂ emissions, integrating AI into security systems can become a valuable tool for achieving sustainability and energy-efficiency objectives.
Regulation is struggling to keep pace with technology
While facial recognition and intelligent analytics applications continue to expand rapidly, regulation is evolving at a much slower pace.
London's Metropolitan Police, for example, reportedly scanned around one million faces during 2025 and has already announced plans to permanently install live facial recognition cameras across South London.
At the same time, the European Union's Artificial Intelligence Act (EU AI Act) began enforcing its first provisions on prohibited AI practices and AI literacy in February 2025, establishing different risk categories for AI systems and the corresponding legal obligations.
In this context, embedding privacy by design and ensuring regulatory compliance are no longer simply legal requirements; they are becoming key competitive differentiators for systems integrators and installers.
The hidden cost of artificial intelligence
Although the technological benefits are undeniable, AI deployment also introduces significant economic challenges.
Average monthly enterprise spending on AI reached USD 62,964 in 2024 and is expected to increase to USD 85,521 per month in 2025, representing a 36% rise. Furthermore, nearly half of all organisations expect to invest more than USD 100,000 per month in AI.
Unlike traditional hardware or software business models, AI costs are driven by continuous resource consumption. Permanent monitoring systems, video analytics and compliance platforms continuously process information, generating costs even when no user is actively interacting with the system.
For organisations operating AI-powered video surveillance platforms, this is already translating into direct pressure on profit margins.
Several studies estimate that between 30% and 50% of AI-related cloud spending is lost to underutilised resources, oversized infrastructure and poorly optimised workloads. Consumption-based pricing models further compound the issue, with 78% of IT decision-makers reporting unexpected SaaS costs linked to AI pricing.
A business model built for AI
Against this backdrop, Anna Acopian identifies three strategic approaches for businesses in the sector: integrating AI analytics as a value-added service on existing contracts; developing AI-based managed services; or specialising in the integration of third-party AI platforms without assuming the operational costs associated with AI models.
Regardless of the strategy chosen, the recommendation remains the same: plan AI costs with the same level of detail applied to physical infrastructure deployment, defining unit economics, expected utilisation and contractual safeguards from the outset to ensure long-term project profitability.
According to Acopian, organisations that combine a robust commercial model with a clear compliance strategy and accurate cost control will be best positioned to unlock the full potential of artificial intelligence.
AI technology is already here. The real challenge now is adapting business models to generate sustainable long-term value from it.
