How artificial intelligence has gone mainstream for security and risk analysis
By Adrian Kotowski
When it comes to security and risk assessments, the use of artificial intelligence has gone mainstream, improving results for analysts when assessing data.
In an industry that deals with life and death scenarios, the adoption of artificial intelligence (AI) to support physical security assessments and activities is no longer a novelty or emerging differentiator. It’s a baseline capability.
While other industries are still evaluating how best to use AI, in the security and risk sector the conversation has moved from whether AI should be adopted, to how it's governed, integrated and operationalised to deliver defensible outcomes.
The most compelling solutions use AI to reduce noise, surface weak signals and prioritise human attention, rather than overwhelm operators with alerts, or replace professional judgment. This reflects the growing recognition that modern security operations suffer not from a lack of data, but from an excess of it, and that poorly implemented automation can increase risk rather than reduce it.
Security decisions have real-world consequences for safety, civil liberties, and organisational reputation. As a result, there is broad acceptance that accountability cannot be automated away. Even the most enthusiastic AI promotor is not considering AI as a replacement for human judgement when assessing risk.
Instead, data integration is emerging as the point where automation meets responsibility, with 'human-in-the-loop' design seen not as a constraint, but as a core strength.
While AI has the capability to supercharge analytical engines, identify anomalies, patterns and correlations, humans retain decision authority, contextual understanding, and the ethical and legal accountability for risk and security assessment.
This has meant that 'human-in-the-loop' design has become non-negotiable, driven as much by trust, liability, and regulatory exposure as by ethics.
How it’s being adopted
While other industries are seeing notable job losses due to AI replacement, particularly in technology, our industry is taking an integration over replacement approach.
For those responsible for purchasing decision making, solutions supporting auditability, decision traceability, and post-incident review are viewed more favourably than those focused solely on speed or autonomy, providing an extra layer of protection for organisations in what is an increasingly litigious space.
Rather than advocating wholesale rip-and-replace approaches, mature vendors are emphasising how to layer AI capabilities onto existing environments, including established VMS platforms, access control systems, and command-and-control frameworks. This approach also reflects the commercial reality, with organisations operating in complex, heterogeneous environments, and unwilling to discard existing investments.
A strong trend we are seeing across the sector is a move towards genuine data-led security models. This has seen AI-enabled video analytics, access control, radar, LiDAR, and other sensor technologies presented as components of integrated decision-support ecosystems, rather than stand-alone tools.
End users, integrators, and security leaders are examining vendors on substance rather than novelty. Discussions are focused on practical operational questions including:
- how systems manage false positives and negatives
- their impact on frontline staff and control room operators
- ease of integration with legacy infrastructure
- and long-term costs of ownership, maintenance, and data governance.
This reflects a disciplined procurement mindset by the risk assessment and security industry that is all too cognisant of past experiences, where promised transformation failed to deliver sustained value.
An integrated future
The approach taken by the security and risk industry to AI is pragmatic and reassuring.
AI may flag, correlate, and rank risk, but people still decide. Technology is being adopted as a force multiplier, enhancing situational awareness, reducing cognitive load, and supporting faster, better-informed decisions, rather than as a substitute for professional expertise. This framing aligns with the realities of security and emergency management, where context, judgment, and experience remain irreplaceable.
With physical security across the Asian-Pacific region becoming increasingly data-led, integration-driven, and human-accountable, organisations are seeking systems that can fuse diverse data sources into coherent operational pictures, while ensuring responsibility and ethics remain firmly in human hands.
The challenge is not adopting AI, but embedding it intelligently within governance frameworks, operational processes, and workforce models that build trust rather than undermine it.
AI has found its place in the physical security profession - not as the hero of the story - but as a capable and increasingly indispensable enabler.
Contact Adrian Kotowski
ERS’s Emergency Options Centre (EOC) is using AI to assist with risk and intelligence assessments for our clients.
Talk to us today about how we can improve your organisational oversight and risk management.
Find us on socials
© 2026 Executive Risk Solutions. All rights reserved.
Executive Risk Solutions Head Office
