8,900 Unmonitored Offenders: What the NAO Report Reveals About the Real Cost of Scaling Electronic Monitoring Too Fast

The UK National Audit Office dropped a 63-page report on 10 July 2026 that should worry every corrections agency in the world considering rapid electronic monitoring expansion. The verdict, delivered by Comptroller and Auditor General Gareth Davies to Parliament: the system is “not working as intended, creating public protection risks.”

That finding alone would warrant attention. What makes this report essential reading is the scale of what went wrong — and the uncomfortable parallels it draws with EM expansion programmes in the United States, Australia, and across Europe.

What Did the NAO Actually Find?

England and Wales have doubled their electronically monitored population from 13,400 in January 2021 to 28,700 by March 2026. The expansion was deliberate policy: more community sentences, wider use of court bail tagging, post-release curfews for prison leavers. The Sentencing Act 2026 will push that number higher still, with an estimated 22,000 additional individuals tagged per year from 2027.

The problem is that nobody can say with certainty how many of those 28,700 people are actually being monitored. As of March 2026, HM Prison and Probation Service (HMPPS) was reviewing 8,900 cases — 24% of the tagged population — where individuals had an active monitoring order but no functioning tag. The Ministry of Justice disputes the figure, placing the true count at around 5,450. Either number represents a system in serious trouble.

How Does a Quarter of the Tagged Population End Up Unmonitored?

The NAO report traces the failure to multiple interlocking causes, and this is where the analysis gets instructive for agencies worldwide.

In October 2023, the Ministry of Justice awarded new contracts to Serco (for tag fitting and monitoring) and Allied Universal Electronic Monitoring (for tag hardware and the monitoring platform). The transition went live in May 2024. Within months, performance collapsed. The backlog of visits to fit, check, or remove tags surged from 1,200 in May 2024 to a peak of 7,000 in October 2024 — a sixfold increase in five months.

By November 2024, Serco had driven the backlog back below 400, a genuine operational recovery. But the NAO’s point is that supplier performance improvements are necessary but insufficient. The deeper failures are systemic.

Why Did Tag Fitting Fail at Such Scale?

Here is one of the report’s most damning statistics: in February 2026, although Serco met its 95% timeliness target for tag-fitting visits, it successfully fitted tags on only 62% of the individuals it visited within its standard two attempts. For the period January to December 2025, that figure was even worse — just 45%.

Think about what that means operationally. A field officer drives to a location, the individual is not home, refuses to comply, or the address is wrong. The officer tries again. Still unsuccessful. The system records the visit as “timely” because the officer showed up on schedule. But the person walks away untagged.

The root causes cited in the report include errors in tagging orders from courts and prisons, incorrect address data, and a low success rate for locating individuals at their registered addresses. These are data-quality problems — and they point to a system where the information flowing between courts, prisons, probation, and the tagging contractor is fundamentally unreliable.

What Happens After a Breach Is Detected?

Even when tags are fitted and working, the NAO found that breaches often go without meaningful response. Serco’s contract does not require it to notify officials of a breach until 29 to 53 hours after it occurs. In a domestic violence case, that timeline can be the difference between intervention and tragedy.

When breach notifications do reach probation officers and police, the system generates such a high volume of alerts — many of them low-priority or triggered by technical issues rather than genuine violations — that front-line staff cannot distinguish urgent cases from noise. Where outcomes are recorded, nearly half of all breach notifications result in “no further action.” Both serious and minor breaches are often processed through the same workflow, with no systematic triage.

This is a pattern that corrections professionals in the United States will recognise immediately. High alert volumes paired with inadequate triage are the single most common failure mode in EM programmes worldwide, and they consistently degrade both officer trust in the technology and public safety outcomes.

The Probation Staffing Crisis Behind the Numbers

The NAO report does not treat EM as an isolated technology problem. It positions the tagging system within a wider probation and corrections ecosystem that is severely under-resourced. As of March 2026, HMPPS estimated a shortfall of approximately 2,200 probation officers. Even with aggressive recruitment — the government says it will reduce the shortfall to 1,500 by September 2026 — the gap remains enormous.

One probation officer quoted anonymously by the BBC captured the operational reality: “There aren’t enough of us, and we have no idea how the government is going to make it work so that nobody is at risk. Because something bad will happen, someone who is dangerous and isn’t monitored will kill someone.”

That is not speculation. It is the professional judgment of a front-line officer watching demand scale while capacity remains flat. The NAO projects that the Sentencing Act’s expansion will generate an estimated 4,200 additional bail breach cases requiring police response and custody each year — on top of existing caseloads.

Why Expanding a Broken System Makes It Worse, Not Better

The NAO’s central conclusion is worth quoting in full: “Expanding a system that is not yet working as intended heightens risks to future performance and, ultimately, to public protection.”

Sir Geoffrey Clifton-Brown, chair of the House of Commons Public Accounts Committee, put it more bluntly: “Most concerningly of all, the government does not know for certain how many people who should be tagged are being left unmonitored, nor does it have the information or capacity to respond quickly to breaches, resulting in unknown risks to public safety.”

The government’s response has been to invest: £100 million in EM improvements, £175 million for expansion over 2026–2029, £700 million in probation, and 2,300 trainee probation officers recruited over two years. These are serious commitments. But Pia Sinha, chief executive of the Prison Reform Trust, identified the strategic risk: “As ministers look for ways to manage prison overcrowding, there is a risk of seeing electronic monitoring as a panacea. Technology can support better outcomes, but it cannot substitute for the skilled probation staff, effective community services and tailored support that help people move away from crime.”

What This Report Means for EM Procurement Worldwide

The UK experience under the NAO’s microscope is not a uniquely British failure. It is a case study in what happens when EM expansion outpaces three critical capabilities: data infrastructure, workforce capacity, and breach-response protocols.

For agencies in the United States evaluating EM expansion — particularly those preparing for increased community supervision under state sentencing reform — the NAO report offers five actionable lessons:

1. Tag-fitting success rate matters more than visit timeliness. A contractor that shows up on time but fits only 62% of ordered tags is not delivering a monitoring service. Agencies should track installation completion rates, not just scheduling metrics. RFPs should specify minimum first-visit success rates and define consequences for persistent fitting failures.

2. Breach-notification timelines must be measured in minutes, not days. A 29-to-53-hour window between breach detection and official notification is not a monitoring system — it is a logging system. Modern GPS ankle monitors with adaptive multi-mode connectivity (BLE/WiFi/LTE) can deliver real-time alerts. Agencies should demand contractual SLAs for breach notification measured in minutes, with escalation protocols based on offender risk level.

3. Alert triage is a design problem, not a staffing problem. When half of all breach notifications result in “no further action,” the system is generating noise that degrades officer response to genuine threats. Device-level intelligence — GPS ankle monitors that distinguish between technical disconnections and genuine tamper events through fiber-optic detection, for example — can reduce false-positive alert volumes by orders of magnitude.

4. Data quality between ordering and fitting must be treated as critical infrastructure. The UK’s 38% fitting-failure rate is driven substantially by incorrect orders, wrong addresses, and information gaps between courts and field officers. Integrated case management systems with real-time data synchronisation between courts, corrections, and monitoring contractors are not a luxury — they are a prerequisite for any EM programme serving more than a few thousand individuals.

5. Workforce planning must precede technology expansion, not follow it. The UK expanded its tagged population by 115% over five years while running a 2,200-officer probation shortfall. Technology does not supervise offenders — officers do. Every EM expansion plan should include a staffing analysis that accounts for the incremental caseload each additional tagged individual creates.

The Bigger Picture: EM as Prison Alternative Demands System-Wide Reliability

The UK is not alone in turning to electronic monitoring as a pressure valve for prison overcrowding. The United States, with state-level sentencing reform accelerating in Texas, Florida, Ohio, and California, faces identical scaling challenges. Australia’s recent expansion of home detention programmes carries the same risks. France, the Netherlands, and Germany have all expanded EM populations since 2020.

The NAO report should be required reading for every EM programme director, procurement officer, and policy maker considering rapid expansion. Its findings are not arguments against electronic monitoring — they are arguments for doing it right. The technology exists to deliver reliable, real-time, low-false-alarm monitoring at scale. But technology deployed into a system with fragmented governance, unreliable data, and under-resourced response capacity will fail, regardless of how sophisticated the hardware.

The NAO’s Gareth Davies summarised it with characteristic understatement: “Improvements are required to ensure that those who should be monitored are monitored and that breaches are responded to effectively.” The unstated implication is harder: if a government cannot ensure basic monitoring of 28,700 people, adding 22,000 more per year is not ambition — it is risk.

Sources

• National Audit Office. “Electronic monitoring: improving resilience to meet increasing demand.” HC 266, Session 2025-26. Published 10 July 2026.
• National Audit Office. Press Release: “Electronic monitoring system is not working as intended, creating public protection risks.” 10 July 2026.
• BBC News. “Thousands of offenders not wearing electronic tags, report says.” 10 July 2026.
• The Guardian. “Expansion of electronic tagging in England and Wales will put public at risk, watchdog warns.” 10 July 2026.
• The Independent. “Thousands of offenders are not wearing tags, report warns.” 10 July 2026.
• Prison Reform Trust. Response from Chief Executive Pia Sinha, 10 July 2026.