The Independent Sentencing Review, chaired by David Gauke, was published in May 2025, and brought into the open – more or less – an alarming vision of the Probation Service’s future.
Implementation, under the rubric of ‘Plan for Change’, began apace, including, in May itself, the first roundtable discussion with the corporate tech sector about their expected contributions to justice innovation.
In June the Ministry of Justice (MoJ) announced £700m “new money” to support the Probation Service up to the 2028-29, although without any initial clarity as to how exactly it would be spent. Precise priorities had not then been set: The ‘Our Future Probation Service’ project, established in February 2025 to improve performance and reduce workloads by 25 per cent by 2027, was still working on them.
In his July McWilliam’s lecture, Lord Timpson, the Prisons, Probation and Reducing Reoffending Minister, extolled the rehabilitative virtues of the old Probation Service, but was singularly unclear – no less that Gauke himself – as to how these would survive and thrive in a future Service focussed so explicitly on punishment and surveillance technology.
The MoJ made no official response to the Gauke Review, but published a Sentencing Bill in September 2025 to take forward what was, in effect, their joint agenda. A month afterwards the MoJ launched ‘Justice Transcribe’ into the Probation Service, a time-saving AI tool for speedily summarising and transcribing conversations with supervisees, in which massive hopes were being invested as a contribution towards resolving the crises of capacity, staffing and performance in the Service.
Keeping a close eye
The Sentencing Bill itself concentrated on more directly punitive technologies, which on the face of it contribute nothing towards resolving probation’s crises. It promised 30 per cent increase in the use of electronic monitoring (EM) – “the biggest expansion of tagging since the adoption of curfew tags in 1999”.
Numbers on EM were growing – 28,000 people were tagged at the end of 2025 – but achieving the MoJ’s target of 22,000 more (by an unspecified date) was a tall order. One contribution towards it was “a presumption that all individuals leaving custody will be electronically monitored for the period they would otherwise have been in custody... This will ensure probation can keep a close eye on thousands more individuals”.
“Keeping a close eye” was becoming a common trope in MoJ discourse on the future of the Service. Speaking of a new, four-site pilot scheme announced in September 2025, which would use remote check-in technology on offenders’ phones, and AI to confirm their identity, possibly in conjunction with GPS tracking, Lord Timpson said:
This new pilot keeps the watchful eye of our probation officers on these offenders wherever they are, helping catapult our analogue justice system into a new digital age
The emerging sense that the old Probation Service was being reconfigured as a punitive-surveillance agency was strongly affirmed when former Lord Chancellor Shabana Mahmood, looking back, declared her real intentions (£):
When I was in Justice, my ultimate vision for that part of the criminal justice system was to achieve, by means of AI and technology, what Jeremy Bentham tried to do with his Panopticon. That is that the eyes of the state can be on you at all times.
Concerns over direction of travel
Over the twelve months following publication of the Gauke Review, the Probation Inspectorate (April 2025); the National Audit Office; the House of Lords Justice and Home Affairs Committee (PDF); and the House of Commons Committee on Public Accounts all published reports concerned about the state of the Probation Service and the direction of travel in which the MoJ was taking it.
The Inspectorate lamented continuing failings in leadership, staffing and services, inadequate material infrastructure (poor quality buildings) and outdated ICT systems. It noted cryptically that “there will need to be significant change to ensure sufficient capacity within the Probation Service to meet operational demand and improve the quality of services” without indicating what that change would be.
The recent HM Prison and Probation Service response (HMPPS) to the Inspectorate’s criticisms stated explicitly that the time-saving digital tools being introduced into the Service were indeed that change, the key to how capacity and quality of service was to be improved. It becomes increasingly difficult to resist the thought that the MoJ is using a human crisis in the Probation Service – one it has no interest in solving on its own terms – to accelerate its transformation into a punitive-surveillant agency.
The House of Lords Justice and Home Affairs Committee “short enquiry” into EM was concerned with making EM a presumptive post-release measure, and – in an as yet unspecified way – “integrating” it into the Probation Service. The “blanket approach to tagging most prison leavers, regardless of crime and circumstances” troubled the Lords because it seemingly “diminishes the role of effective, targeted Probation interventions, and risks creating an unethical system that is overly punitive and disproportionate”.
Just because the MoJ had a legal justification for doing this, said the Lords, did not mean they should. EM had a place, they agreed, but it should not “become a proxy for effective probation work”, and they worried that some of the £700m notionally earmarked for probation would be spent in part on EM.
Wanting to forestall the MoJ’s perceived over-investment in EM, and restore confidence in rehabilitation, the Lords asked the MoJ to revise its most recent EM Strategy (2022), believing that any balanced and evidence-based assessment of EM and the Probation Service’s respective merits would favour the latter. The Lords were somewhat “retrotopian” here, not realising that within the new tech-driven paradigm in which policy on probation is now being taken forward, the MoJ regarded the evidence-base on which the old Probation Service’s authority had once rested, as a little passé.
The Lords’ concerns about EM were sadly not matched in their stance on AI. They did not demur when the MoJ spoke of “the potential for AI to revolutionise our approach” or of “maximising data use” to improve EM. They fell for the simple efficiency argument, that AI would lift “some of the burdens from probation practitioners’ shoulders so they can concentrate their time where it is most valuable.
They seem to have taken some cues from the Confederation of European Probation’s optimism optimism about AI tools, and accepted that any challenges they might pose for Probation would be risen to. Equally, the Lords may have been seeking a trade-off: go for AI, step back on EM.
The Committee for Public Accounts were more sanguine. Echoing the National Audit Office, which had mostly concentrated on the high-risk tech strategy which ‘Our Future Probation Service’ was pursuing, it too was unconvinced that reckless investment in digitalisation was adequate to resolving the staffing and standards challenges facing the Service.
They feared that the pace at which the HMPPS was planning to introduce them could be counterproductive, and was highly likely to “disrupt services, contribute to poor outcomes and staff stress... the short time-frame carries a high level of risk and the MoJ does not have a strong history of implementing digital change programmes well”.
Tipping point
The four post-Gauke reports on the multiple crises facing the Probation Service – and the way those crises are being used by the MoJ to drive fundamental changes in its character and ethos – have yet to be properly synthesised and discussed. There is as yet no organised resistance to the move towards a punitive-surveillant agency, which is not helped by a clear statement from the MoJ on how far it actually wants it to go.
Reassuring talk about only using AI for efficiency measures like transcription is misleading: even Gauke expected it to go further, writing of using “advanced AI” and expecting this to encompass “AI agents” for planning supervision schedules and, possibly, chatbot-driven dialogue with supervisees. The MoJ’s tight relation with the tech industry – particularly its own contract with OpenAI – bodes ill in this respect. It is in the nature of the AI industry to promote continuous innovation, and to hook users with the self-deprecating guarantee that ‘this is the worst AI you will ever have’.
A tipping point may already have been passed, such that resistance to AI-driven public services is already impossible. The emerging ‘digital rehabilitation and desistance’ movement offers slender hope, because while it is expressly not aligned with a punitive-surveillance agenda, it presupposes that digitalisation will be constrained by the culture and values of the ‘old’ Probation Service, and go so far and no further than these values allow.
Quite apart from the manifest threat to these values, a culture of continuous AI innovation makes ‘so far and no further’ a rather pious hope.
