Emergency services face growing pressure from synthetic hoax calls, rising demand and infrastructure failure. Drawing on a decade in policing, James Johnson explores how modern contact management can strengthen resilience, support fast, defensible decisions and help police, fire and ambulance services keep operating when it matters most.
Author
Written by James Johnson, Business Development Executive, Emergency Services
“Everything you’ve been taught about methodical building clearance is now flipped on its head. This is an emergency search. There is an immediate threat to life.”
Thankfully, this rapid tactical intervention was only a training exercise. Squeezing the trigger only hammered a blank round through the barrel, but the pressure felt real.
Firearms officers go through that door knowing their next split-second decision may save a life or end one. Most of us rarely contemplate that responsibility however, synthetic voices can now increase that pressure by manipulating police, fire and ambulance contact centres into triggering a physical response.
Emergency contact centres must therefore remain resilient in three conditions: when they are fooled by false information, flooded by demand or switched off by infrastructure failure. Technology can strengthen that resilience, but only when it supports sound decisions, tested processes and effective coordination between services.
I understand that challenge from both sides of the radio. Over a ten-year policing career, I answered tens of thousands of emergency calls and deployed to incidents using information passed over the radio.
The technology has changed since I handed in my warrant card and headset in 2022, but the purpose remains the same: answer the call, understand the risk and act correctly.
The foundations of resilience
My time in control rooms taught me that resilience is not a document. It is whether the next call is answered, understood and acted on correctly while other systems are under pressure.
Secure networks, tested processes and fallback procedures keep 999 services available around the clock. The fundamentals have not changed: answer quickly, locate accurately, assess consistently, hand over effectively and continue operating when a system, supplier or building fails.
New technology should strengthen those foundations. It should not make decisions harder to understand or create new dependencies that have not been tested under pressure.
Fooled: when the call is not real
It is 02:14. A 999 call connects to a police control room. The caller gives an address, reports a weapon and describes an immediate threat to life. The call handler follows procedure, the control room assesses the threat… ‘spinning the wheel’ of the National Decision Model, and the force deploys an armed response.
Forty minutes later, another force receives a call using the same words, in the same order, spoken by the same voice. The address is different, but the reported threat is identical.
That pattern is not coincidence. It shows how you can connect the dots between computer code and a carbine, creating potentially lethal consequences in the physical world.
Reported incidents abroad show elements of this threat emerging. US journalists linked a series of swatting calls to a service that appeared to use a consistent synthetic voice. In October, Spanish police arrested two suspects accused of disguising their voices with AI and spoofing caller identification during airport bomb hoaxes.
Synthetic hoax calls are a different kind of problem, because the defence can’t simply be “spot the fake”. Detection tools alone are unlikely to provide a complete defence, and no call handler should be expected to identify a cloned voice by ear at 2am. Nor can the answer be scepticism.
The worst would be a delayed response to a genuine threat.
The question is therefore not whether to deploy. It is how to make a fast, defensible decision while independently testing the information provided. Five principles can help:
- Treat the call as genuine and verify in parallel. Plan the response as normal while checking the caller’s account against information lawfully available to the control room. Verification must support the response, not delay it.
- Share suspicious patterns immediately. A force receiving a suspected synthetic call should be able to alert other control rooms quickly. Repeated wording, voices or reported circumstances may become visible nationally before they are apparent locally.
- Brief decision-makers on the threat and the uncertainty. Grounds for suspecting a hoax should inform command decisions and operational briefings. The potential victim of swatting may be the innocent person behind the door.
- Train for synthetic calls without publishing exploitable indicators. Staff need realistic scenarios and the confidence to escalate concerns, but services should avoid publicly identifying the specific indicators on which they rely.
- Preserve evidence and close the intelligence loop. Retain relevant material, share intelligence through appropriate channels and work with telecommunications providers where tracing or spoofing is involved.
None of this is police-only. A credible firearms call can stand up ambulance and fire and rescue resources, and a synthetic caller could just as easily report people trapped in a burning building. The Joint Emergency Services Interoperability Principles (JESIP) should also inform how agencies share intelligence about suspected hoaxes.
The operational lesson is that a convincing voice cannot be treated as proof of identity, but uncertainty must not become a reason to delay a genuine emergency response.
Flooded: when demand exceeds capacity
Contact centres must also remain resilient when genuine demand arrives faster than services can process it. This is not only a staffing problem. Concurrent calls, repeat contact and movement between telephone and digital channels can make it harder to identify which incidents need the fastest intervention.
The Grenfell Tower Inquiry’s control-room recommendations show why this matters. Fire and rescue services must be able to manage large numbers of concurrent calls while maintaining effective communication with incident commanders. Ambulance contact centres face their own demand pressures, while police forces continue to balance emergency calls, 101 demand and growing digital contact.
Technology can help when it addresses a defined operational problem. West Yorkshire Police’s Post Call Analysis seeks insight from historic control-room calls, including patterns of repeat contact. The Bobbi assistant is intended to support non-emergency contact, while Wiltshire’s abandoned-call text messages provide another route back to the public. These examples use technology to understand demand, reduce friction or redirect contact. They do not remove the need for trained staff and tested surge arrangements.
Technology improves resilience when it helps services identify demand, remove avoidable contact and protect staff capacity for the incidents that need human judgement.
Switched off: when infrastructure fails
Contact management can also fail even when staff make the right decisions. A network outage, loss of power, unavailable building or supplier failure can interrupt access to emergency services.
The transition from analogue to digital telephone services changes part of that risk. Digital lines do not provide their own power in the same way as traditional lines, which matters for people who depend on landlines or telecare equipment. Emergency services are also managing the transition from Airwave to the Emergency Services Network while maintaining operational continuity.
The disruption to UK 999 services on 25 June 2023 demonstrated the consequences of infrastructure failure.¹ Resilience therefore requires more than an uptime commitment. Services need tested fallback routes, rehearsed degraded-mode working, pre-agreed mutual aid and contracts that define how suppliers will recover essential services.
¹ Department for Science, Innovation and Technology & Cabinet Office, Public Emergency Call Service disruption, Sunday 25 June 2023: post-incident review (21 March 2024)
What industry owes the sector
Emergency services should judge suppliers on how their services perform under operational pressure, not only on how they perform in demonstrations.
Suppliers should demonstrate how their services respond when contact centres receive deceptive information, demand exceeds capacity or critical infrastructure fails.
That means designing for failure, supporting open standards, making AI outputs explainable and enabling data to move safely between forces and services. No single product will solve synthetic calls or guarantee resilience. The defence is layered: people, process, data and technology working together.
Conclusion
Modernising contact management is not simply a technology programme. It is an operational-resilience challenge. Emergency services must recognise uncertainty, verify information without delaying necessary action, share emerging threats and continue operating when systems or suppliers fail.
Industry’s role is to strengthen that system, not sell a single answer to it.
I will be discussing these issues at the techUK resilient communications roundtable on 23 November. I would particularly welcome views from people working in control rooms: what helps teams make fast, defensible decisions when a call does not sit right? And where should national coordination begin: shared alerts, shared standards or shared platforms?