AI is not just driving the convergence of physical security and cybersecurity—it has totally blurred the distinction between the two by enabling an integrated approach to protection and access that spans traditional systems and adds new capabilities to improve security and reduce risk.

At the same time, AI has effectively merged physical and cybersecurity defenses into a single attack surface. Today’s AI-powered threats target both digital and physical access controls with autonomously chained tactics ranging from deepfaked biometric credentials to malicious drone surveillance to socially engineered breaches of core operational systems. 

Exposed organizations across industries must rise to meet these emerging threats. But a strategic solution goes beyond just grafting AI onto existing systems. The rearchitecting of siloed security frameworks to integrate AI-enabled threat analysis and automated response has become an urgent necessity to safeguard buildings, people, sensitive data, and business services.

What does a merged physical/cybersecurity environment look like? What capabilities can it provide to improve security effectiveness and counter emerging threats? This article gives business and technical leaders a comprehensive overview.

Key takeaways

• The barrier between digital and physical security has already dissolved as both threats and defensive solutions merge and overlap. 
• Integrating AI surveillance tools with cybersecurity controls can accelerate threat response and streamline operational efficiency. 
• AI can enhance the speed and effectiveness of human security staff by improving event detection, accelerating response time, and reducing operational costs.
• AI-powered convered security can improve protection with capabilities like integrated incident response, faster incident detection, proactive alerting on a wider range of safety and security factors, and improved anomaly detection. 
• Converged security portfolios create complex governance challenges, including ethical issues or privacy violations from intensive surveillance, bias concerns in physical recognition algorithms, and the potential for catastrophic damage if unified security control is hijacked.

What does a modern physical/cybersecurity environment look like?

The boundary between physical and digital security has already dissolved. When physical security solutions connect to digital networks, they become part of the cyber-attack surface. Likewise, without robust physical controls to protect IT assets from direct attacks, cybersecurity infrastructure can be readily bypassed. 

AI-enabled integration and automation of today’s standalone and bandwidth constrained cyber and physical security infrastructure and monitoring/analytics processes is the only way to achieve a preemptive defensive posture to counter AI-powered attacks.

Rather than just recording events for delayed or reactive analysis, converged security teams need real-time, proactive situational awareness and incident detection. AI systems can process massive data streams, including input from cameras, access control solutions, OT systems, and IoT devices, to instantly detect, predict, and respond to possible threats.

A popular emerging use case is to integrate physical access control data, such as surveillance cameras and badge reader records, with network login data to compare a user’s physical location with their digitally tracked activities.

Emerging use cases for converged security technology include:

• More intelligent access control. AI systems can link facial recognition or other biometrics with multifactor authentication software to control access to restricted digital or physical assets.
• Automated threat detection/prediction. AI-powered security platforms can leverage machine learning across large data volumes, such as logs from physical and digital security controls, to predict impending physical or network breaches before the attack can be completed.
• Intelligent buildings. AI can integrate environmental sensors with HVAC, video surveillance, and network cybersecurity to block attacks on connected devices and physical assets at the same time.

How is AI changing physical security?

AI is revolutionizing physical security operations across private sector, critical infrastructure, and public safety organizations as it revolutionizes access control, surveillance, and threat detection systems. AI can greatly improve the proactivity and effectiveness of human security staff by enhancing event detection, accelerating response time, and cutting operational costs.

AI-supported physical security programs can be overall more effective and reliable than pre-AI solutions at preventing incidents and minimizing their impacts. AI also reduces the need for human monitoring, helping to save costs and free security staff for higher-value tasks. 

Top capabilities that AI currently brings to physical security include:

• Enhanced video surveillance and monitoring.
  AI adds enhanced object recognition, behavior analysis, and real-time anomaly detection to video surveillance. Camera manufacturers are now embedding AI and machine learning into their cameras to provide onboard facial recognition, license plate recognition, and criminal activity detection. AI-powered video systems go beyond traditional movement detection to identify threats with greater speed and accuracy, including unauthorized intrusions, suspicious or hostile behavior, and abandoned objects. AI also enables law enforcement to identify specific people or vehicles in real-time.  

• Improved access controls.
  AI elevates physical access control and replaces cards and passwords/PINs with stronger protections like advanced biometric authentication (including voice, facial, fingerprint, and iris recognition), automated access policy enforcement, and automated security response. AI systems can make real-time access decisions, reducing the need for continuous human monitoring. AI can also evolve with changing security threats, including recognizing and alerting on unusual user behavior. 
• Proactive incident prediction and prevention.
  AI can analyze huge data volumes to identify patterns, predict possible security incidents, and alert staff. This reduces risk by enabling security teams—or the AI itself—to take preventive measures before problems occur. For example, AI can monitor sports events, concerts, and other large gatherings to predict impending crowd surges or stampedes so security personnel can proactively deescalate the situation. AI can even flag security vulnerabilities based on historical data. 
• More capable and reliable security drones and robots.
  AI-powered autonomous systems can now patrol specific areas, analyze their surroundings in real-time for suspicious activity, and act on possible threats using alerts or voice. Drone surveillance opens up a broader view of secured areas, especially around large facilities or remote sites that are challenging for humans to cover. Some equipment features environmental sensors that can detect hazardous materials or scan for temperature variations.
• More effective emergency response.
  Because of their ability to rapidly process multiple input streams and direct multiple responses in real-time, AI can be extremely valuable in natural disasters, fires, and other emergencies. AI can automate wide-scale alerts, provide a real-time situational view to responders and authorities, and otherwise help reduce human response times, improve response effectiveness, and blunt crisis impacts. 

How can physical/cybersecurity convergence improve protection?

As it does with cybersecurity alone, AI can help identify vulnerabilities and eliminate risks in real-time across converged security infrastructure. By integrating these formerly separate security aspects, organizations gain a holistic view of security risks with overarching policies, fewer blind spots, better cross-team communication, unified incident detection and response, and improved business continuity.  

Converged security platforms use AI capabilities like behavior-based analytics and enhanced threat detection to accurately identify a wider range of threats, such as unauthorized physical access, unusual logins to software systems, and suspected data breaches.

Use cases where physical/cybersecurity convergence can help organizations safeguard people, premises, and data include:

• Integrated incident response.
  Processing digital and physical security data together helps identify incidents that include both vectors. For example, a common way for hackers to breach networks is by physically installing a rogue wireless access point, such as a hotspot, plug-in access point, or router. This requires physical proximity to a wireless network’s coverage area. AI-enabled technology can detect rogue access points and trigger cameras to record the physical area and analyze users’ behavior to catch the hacker in the act. 
• Faster incident detection.
  Converging cyber/physical security data on a common dashboard improves threat visibility across the board. For example, staff can lock down data sources and building areas at the same time if a physical attack on server hardware is detected. 
• Proactively alerting on a wider range of issues.
  AI systems can integrate with environmental sensors and other IoT devices to alert on a wider range of safety and security parameters like body heat mapping, air temperature, air quality, and noise levels. For example, AI can process video feeds to monitor patient safety in hospitals, such as alerting staff when a patient appears to have fallen or moves toward the edge of their bed. 
• Detecting anomalies or unusual behavior.
  AI-supported continuous monitoring using cameras, sensors, or intrusion detection devices are more powerful when paired with human experience and insight. AI can alert security staff if a window is left open that is typically closed, for instance. AI can also notify a human in the event of suspected vandalism, trespassing, loitering, or theft.

What percentage of organizations have unified their security platforms today? According to Jeffrey Friedman, CEO at Fortifye, that number is still very small but has the potential to rise quickly.

“I feel like I’m on the edge working on this with clients and showing them the tool sets and that they can actually accomplish these kinds of goals,” says Jeffrey. “It’s not really that hard, because the data is there. The convergence of it all is available.”

What are governance concerns with converged, AI-powered security?

Converged, AI-powered security solutions can create new governance and accountability risks. These include algorithmic biases in physical recognition protocols, privacy concerns around intensive surveillance, and potential exposure to overwhelming negative impacts if the unified security platform is compromised.

Examples of new risks with converged, AI-powered security services include:

• Discrimination against minority groups due to misidentification with facial recognition algorithms or other access controls. Rigorous auditing of AI behavior is key to preventing discrimination in physical access restrictions or false arrests/detainments.
• Privacy concerns due to mass surveillance or tracking while harvesting digital data (e.g., location data) alongside biometrics, CCTV footage, and other physical security data. This can create privacy risks for individuals as well as compliance violations. 
• Ethical questions around who owns these new converged data stores and how they can monetarily benefit from them. 
• The potential for a single point of failure where a unified AI system controls both physical and cybersecurity controls. In this scenario, a successful attack on the control surface could put both an organization’s facility and its most sensitive data at risk. Effective governance requires human-in-the-loop controls as a fail-safe mechanism.

What’s next?

For more guidance on this topic, listen to Episode 159 of The Virtual CISO Podcast with guest Jeffrey Friedman, CEO at Fortifye.