Guest article by John Oswald, Managing Director, Public Safety, Accenture
The UK justice system does not have one communications problem. It has hundreds of them. This is not at all a criticism of the people doing the work. Quite the opposite. I know many people in justice communications. Across HMCTS, HMPPS and the wider ecosystem, there are excellent teams working across digital, PR, campaigns, strategic communications, operational updates, internal engagement and service communications. Much of their work is high-pressure, high-stakes and largely invisible to the public, and yet is a huge component of building and maintaining trust in the system.
But communications in justice often inherits the shape of the system around it, which as we know is fragmented. That’s not news by now, of course, and we’ve seen Sir Brian Leveson articulate it at every level.
Communication in justice is not decoration. It is not the final layer applied to a policy, a service, a court process or a campaign once the “real” work has been done. In justice, communication often is the real work.
It affects whether people understand what is happening to them. Whether they attend. Whether they comply. Whether they trust the next step. Whether they feel believed, respected or degraded. Whether they give up.
A court update, a victim notification, a prison-family communication, a probation instruction, a GOV.UK page, a campaign, a press line and a public statement are not the same thing. They have different channels, constraints and audiences. But they should be animated by the same duties.
Accuracy, because justice is a legal system, not a customer service brand.
Humanity, because people are often frightened, angry, ashamed, traumatised or simply overwhelmed.
Dignity, because the state is exercising power, and requires trust.
The backdrop to all this is Sir Brian Leveson’s review of the criminal courts. It is, in every sense, a system diagnosis. It describes delay, complexity, strain and fragmentation. It shows how problems upstream and downstream of the courts compound each other. It makes clear that criminal justice is not experienced as a tidy process flow. Far from it. It’s often experienced as waiting. As silence. As handoffs. As uncertainty. As letters and portals and phone calls and hearings and forms. As a sequence of moments in which people may or may not understand what is being asked of them.
And so, if the system is fragmented, the communications will be fragmented too, unless we design against that fragmentation.
That is where citizen-centred communications becomes much more than “plain English”, important though that is.
For me, citizen-centred communications in justice means designing every communication around five questions.
This applies whether we are talking about a national campaign, a media response, a GOV.UK page, a victim update, a court notification, a probation instruction, a prison-family communication or a service email.
Different channels. Same human being. Same system. Same duty.
This is where I think communications and user-centred design need to move closer together.
Too often, they operate as adjacent disciplines. User-centred design asks what someone needs to do, understand or feel confident about, and usually within the context of a digital platform or service. Communications asks what an organisation needs to explain, influence, protect or land.
In justice (as indeed in many organisations), those questions should overlap a lot more.
The same person who encounters a campaign may later receive a service update. The same victim who reads a public commitment may later sit waiting for a case update. The same family member who sees a message about rehabilitation may later try to navigate contact with someone in prison. The same defendant who is told the system is fair may later receive dense, frightening, procedural correspondence.
Tone can flex by context, and indeed should. A public campaign is not a court order. A media line is not a victim update. A prison notice is not a service design pattern. But the institution should not feel like a different institution every time. To put it another way, you can’t build public trust in the abstract, then spend it away in the interaction. Not unlike how the idea of ‘brand’ behaves in companies – it’s still a challenge for many to balance the brand promise with the service experience of the organisation itself.
To echo my wonderful colleagues Morna Spence and Emma Feggetter: Morna wrote about the need to move beyond efficiency and think about a justice system that works as one system. Emma asked what a digital-first justice system actually looks like: not digital gloss, but better handoffs, captured-once information, proactive updates and services organised around the whole journey.
My question builds on these: what is the communications equivalent of that?
I think it starts with accepting that citizen-centred communications is not simply about external campaigns, though they matter. It is not simply about PR, though public debate matters. And it is not simply about the letters, notifications and updates sent to people inside the system, though those may be the most important communications of all.
It is about coherence between them. If a service presents itself publicly in one tone, but speaks to people in a live case in another, trust is weakened. If one agency has insight into what victims need at a certain stage, but another team starts from scratch, the system learns too slowly.
If every campaign brief, supplier brief, content brief or service update begins as a blank page, we waste scarce energy and lose institutional memory.
There are practical ways to do better. For example, a shared justice communications insight layer: a way of pooling what teams already know about victims, witnesses, defendants, prisoners, families, practitioners and the wider public.
This could then extend into journey-based communication principles: not one flat tone of voice, but guidance that recognises where someone is in the system and what they may need at that moment.
Supporting all that would be reusable briefs, content patterns and message frameworks so that agencies, teams and suppliers are not always starting from scratch when producing new material. That also extends into mechanisms for sharing content and evidence across teams, especially where the same audiences are being reached through different channels.
In short, the system as a whole would do well to bridge service design, policy, digital, operational delivery and communications much earlier, as a way of engendering better trust in outcomes overall.
And yes, we should use newer tools intelligently to support all this. There is an obvious efficiency story: drafting variants, summarising insight, checking consistency, improving accessibility, supporting translation, helping teams create better briefs and reducing duplication. But justice needs a more careful AI story than “faster content production”. The risks are real: inaccuracy, bias, synthetic empathy, copyright issues, flattening of voice, loss of human judgement and a public sense that the state is automating its humanity.
The opportunity is not to make justice communications less human: it is to give humans better tools, better memory and better context. That requires governance, standards and accountability. And it requires knowing where AI should not be used, as much as where it can help. In the end, citizen-centred communications is not about making justice sound nicer; It is about making the justice system more intelligible and trustworthy to the people who are caught up in it, affected by it, working within it or trying to understand it from the outside.
Ultimately the way we communicate justice is just as important as the system itself. Let’s treat it as such.