Under constant surveillance? Restriction zones for released offenders present profound legal, ethical & practical difficulties.

The notion of separating offenders from victims through geographical boundaries is a tale as old as time. Historically, offenders were banished into exile to punish and shame. More recently, the law has implemented legislation that strives to manage risk with protection for complainants:

• Non-molestation orders were introduced by the Family Law Act 1996, one of the earliest UK laws allowing courts to prohibit an abuser from harassing or contacting a victim (including exclusion from certain areas).
• Restraining orders have been available since the Protection from Harassment Act 1997.
• Sexual harm prevention orders under the Sexual Offences Act 2003 can exclude offenders from both physical and online spaces.
• Domestic violence protection orders and domestic abuse protection orders have been designed to remove perpetrators from homes and neighbourhoods.
• Antisocial behavioural orders carved out ‘no-go zones’ around town centres and estates.
• At the most extreme, terrorism prevention and investigation measures have enforced relocation altogether.

The key problematic difference with the government’s latest proposal is that it is an inversion of the policies described above, which are all exclusion zones to (justifiably and proportionately) prevent entry. Restriction zones, however, force individuals to remain within one area. This replicates a concept akin to incarceration, which would restrain an individual’s movements even further.

Government promises

The announcement forms part of the government’s Plan for Change to ‘boost protection for victims and make our streets safer’, to allow victims to travel freely without fear of accidental  encounters with those who caused them harm in the first place.

Probation officers will assess risk and consult victims directly in drawing up boundaries. Enforcement will be through extensive GPS tagging, with the government committing £700m by 2029 and recruiting at least 1,300 new probation officers.

What is conspicuously absent is any commitment to consult offenders themselves, or to set out enforcement mechanisms and human rights safeguards.

Given the premise of the proposal, it would not be surprising if the government has also contemplated its potential to alleviate the issue of early release and overcrowding in prisons.

Legal fragility & ethical paradox

The policy has been hailed as a victory by campaigners. However, the finer details have yet to be released to the public and sufficiently scrutinised.

The government announcement promises to ‘curb the freedoms of the most serious sexual and violent offenders’ after their release from prison by ‘locking’ them into specific vicinities.

This does not sound like a sensible and properly considered criminal justice reform. Instead, it reads like a draconian, dystopian episode of science fiction dressed up as victim protection. Arguably, restriction zones constitute punishment that perpetuates beyond prison sentence, and that hinders reintegration and rehabilitation.

The plan is that offenders will be surveilled through technology, some of which is AI-based. We do not have a central AI regulator in the UK to ensure fairness, lawfulness and accuracy.

Without question, victims of (serious) crime deserve protection and peace of mind, but those rights should not come at the expense of liberty, privacy and rehabilitation, which are fundamental legal principles upheld in the UK. They risk violating at least three Articles in the European Convention of Human Rights (ECHR): Art 5 (right to liberty), Art 8 (respect for private and family life) and Art 1 of Protocol 1 (peaceful enjoyment of possessions, including employment and housing).

Terrorism prevention and investigation measures similar to restriction zones were upheld due to the exceptional context of national security (Secretary of State for the Home Department v BM [2012] EWHC 714 (Admin), [2012] All ER (D) 19 (Apr)).

However, it is incongruous, in my view, that the government would be able to justify comparable restrictions in ordinary criminal cases unrelated to terrorism. The proposal appears entirely incompatible with existing case law. In R (on the application of Jalloh) v Secretary of State for the Home Department [2020] UKSC 4, [2020] 3 All ER 449, the court held that confining an immigration detainee to a single address for most of the day amounted to a deprivation of liberty under Art 5 of the ECHR.

Restriction zones, as currently defined, restrict movement at all times, as opposed to just certain times of a day. In practice, these zones are far more onerous than those struck down by the  Supreme Court in Jalloh.

The practical problem

Even if legislation were somehow drafted to overcome the human rights hurdles, it would still be fraught with practical problems that would impact an offender’s access to housing, travel and employment.

The government’s announcement is unclear when it comes to the practical aspects of creating and enforcing restriction zones. However, if offenders were to be concentrated to specific areas,  here is an obvious danger of geographical inequality, with certain communities becoming ‘digital enclaves’, if those communities were to house disproportionate numbers of tagged offenders. This would breed resentment among civilian residents and merely displace risk.

The MoJ states that increased tagging of released offenders and a greater reliance on monitoring technology will help adherence to the restriction zones. However, according to a probation officer quoted by the BBC in the August 2025 article ‘Serious offenders to be kept in restricted zones under plans’, the new zones would put more pressure on the service when it is already struggling to monitor offenders in the community.

The current tagging system is beset by failures. Reports from the BBC, The Guardian and the National Audit Office revealed that prisoners have been released without tags, others wore devices that weren’t activated for days, and that the coordination between prisons, probation and contractors has been poor. False alerts are commonplace as GPS tracking frequently fails indoors, in  dense cities, or under interference. These errors could see lawfully compliant individuals recalled to custody.

A programme to transform the system for electronic tagging was abandoned in 2022 due to unworkable software and supplier delays, wasting tens of millions of pounds of taxpayers’ money. Rather than improving monitoring, the scheme exposed systemic weaknesses and is a cautionary tale for any expansion of tagging into restriction zones.

Technology as state enforcer

Traditionally, the MoJ and probation services relied on human beings to supervise offenders through police surveillance or community reporting. In May 2025, the UK government allocated £700m to expand electronic monitoring increasing tagged individuals from 9,000 to nearly 40,000.

The MoJ is expected to unveil new types of tags capable of detecting illicit drug use by measuring biometric indicators such as heart rate and blood pressure. As a criminal lawyer, it is obvious  that these will raise issues around bodily autonomy, data security and constant surveillance.

We are seeing a shift in responsibility. Enforcement is increasingly reliant on technology such as AI (eg, predictive policing, risk assessments and live facial recognition), GPS bracelets,  geofencing, and smartphone/biometric check-in systems. This outsourcing further entrenches criminal justice in technology. This sounds efficient, neutral and scientific, but in reality, the  technology is built on human choices and assumptions.

Algorithmic and AI-driven risk assessments are used by probation services to detect ‘risky patterns’ before breaches occur. This opaque predictive policing raises important questions about  accuracy and bias. Algorithms cannot reliably distinguish coincidence from intent; they replicate racial and socio-economic disparities using historical data. In other words, if past policing was biased then those biases will be encoded into the system. In February 2025, Amnesty International published a 120-page report on how predictive policing systems are ‘supercharging racism’
in the UK.

“Restriction zones shift the balance between protection & liberty”

In June 2025, MPs attempted to halt automated decision-making, profiling and AI for the purpose of predictive policing through an amendment to the Crime and Policing Bill. Then in August 2025, the Equality and Human Rights Commission was granted permission to intervene in an upcoming judicial review examining whether the Metropolitan Police’s use of live facial recognition technology complies with human rights law.

The Commission acknowledges the potential value to policing but believes that the current policy governing its use is incompatible with Arts 8 (right to privacy), 10 (freedom of expression) and
11 (freedom of assembly and association) of the ECHR.

Justice requires discretion, empathy and context. Relying too heavily on machines risks the erosion of human judgement in favour of perceived efficiency. Moreover, computer systems can make decisions that humans don’t understand. If an algorithm flags something, it may then be difficult to understand and then challenge because the reasoning is hidden behind inexplicable data  and/or intellectual property. This raises issues of opacity and lack of accountability.

Lessons from the US

Most jurisdictions in the US have broad exclusion zones for sex offenders, typically banning them from schools, parks or residential areas.

At least 30 states have laws prohibiting registrants from residing within specified distances (typically 500–2,500ft) of public places such as schools and nurseries where children gather.  Hundreds more local municipalities have created extensive buffer zones that severely limit permitted housing options.

Crucially, however, even at their harshest, these US laws stop short of locking offenders into one zone, and yet,  these exclusion zones have produced damaging consequences. They have left many offenders homeless, clustered in encampments under bridges or on the edges of towns.

The Julia Tuttle Causeway colony in Miami is an infamous example, where stringent restrictions left many registrants homeless and forced to live under a highway overpass. In California, Jessica’s Law enforced a 2,000ft residency restriction and lifelong GPS monitoring for high-risk offenders. In 2015, the California Supreme Court struck this down as unconstitutional
due to its overly broad nature.

Studies show little evidence that such measures reduce reoffending in the US. For example, a study assessing the effectiveness of sex-offender residence restrictions in Michigan and Missouri
(‘Effect and implications of sex offender residence restrictions: evidence from a two state evaluation’, Criminology and Public Policy, vol 13 (2014)) found that if residence restrictions had any effect on recidivism, that impact is small.

As in the UK, GPS tagging has been plagued by false alerts. Numerous legal challenges across different states highlight constitutional problems ranging from vagueness to unlawful banishment and violation of family rights.

The UK’s proposed restriction zones go further, beyond even the most severe US measures. If implemented, they would represent the first attempt to use geofencing not just to exclude, but to
imprison in place.

Conclusion

Restriction zones shift the balance between protection and liberty. They extend punishment beyond custody and risk restraining liberty through machines.

As lawyers, we have a duty to ensure that legislation is targeted, proportionate and legally sound. In my view, these are qualities that the proposals lack.

Victims deserve protection without becoming prisoners of their trauma. But locking offenders in what will amount to ‘digital cages’ is not the answer.

This Article was written by Partner, Henrietta Ronson and was first published in the New Law Journal on 10th October 2025.