Artificial intelligence has spent two years being described as the future of law. In the same week, the Ministry of Justice handed a £3.62 million contract to run the next phase of LawtechUK, and a tribunal confirmed a £60,000 penalty against a Merseyside firm whose client files ended up on the dark web. Read together, the two stories point to the central tension of the UK's legal technology agenda: the government wants AI to become ordinary professional practice, but ordinary practice is only as safe as the data it runs on.
The Money Behind the Machine
The Ministry of Justice has announced that CodeBase, an Edinburgh-based organisation, has won the contract to run the next phase of LawtechUK. The stated aim is to shift artificial intelligence in the UK legal sector from novelty to everyday professional practice. That is a deliberate change of language. For much of the last few years, legal AI has been presented as a showcase, with pilots, demonstrations and headline-grabbing claims. The new framing treats it as infrastructure that solicitors, barristers and legal executives should use routinely, in the same way they use case management systems or e-filing.
What the Contract Signals
Three signals stand out for practitioners:
- Government ownership of adoption. Public money is being directed at helping the profession use technology, not just at building it.
- A move from pilots to practice. The emphasis on everyday use suggests firms will be expected to demonstrate workable, repeatable processes rather than one-off experiments.
- Adoption at scale. Success will be measured by how many practitioners use these tools, which means smaller firms and sole practitioners are part of the intended audience, not an afterthought.
For firms, the practical question is no longer whether AI will be used in legal work but on what terms. Adoption that comes without governance is a liability, and the terms are being set as much by regulators and courts as by technology vendors.
The Cost of Getting Data Wrong
The second story is less glamorous but more immediately instructive. The First-Tier Tribunal has upheld a £60,000 penalty imposed by the Information Commissioner's Office on DPP Law, after hackers compromised its server and published confidential client files on the dark web. According to the Law Gazette's report, the tribunal found the fine proportionate and a crucial deterrent for firms handling sensitive client data.
The tribunal ruled that the Information Commissioner's Office fine was proportionate and served as a crucial deterrent for firms managing sensitive client data.
The significance of this ruling goes beyond one firm. It confirms that regulators can impose meaningful financial penalties for data security failures, and that appeals against those penalties are unlikely to succeed where the failures are clear. Client confidentiality is the foundation of the solicitor-client relationship, and a breach that exposes files publicly is treated as a serious failing regardless of whether the attack itself was criminal.
Where the Two Stories Meet
The link between these developments is direct. AI tools typically depend on large volumes of client data, whether for document review, drafting, research or case analysis. Every new system connected to case files widens the attack surface. A firm that rushes to adopt AI without tightening its security baseline is effectively increasing its exposure to exactly the kind of incident that led to the DPP Law penalty.
Mapping the Risk Against the Opportunity
The table below sets out how the two developments translate into practical priorities for UK firms.
| Area | What LawtechUK Signals | What the DPP Law Case Reveals | Practical Response for Firms |
|---|---|---|---|
| Technology use | AI moving into everyday practice | Systems holding client files are high-value targets | Map every tool that touches client data before rollout |
| Governance | Adoption at scale across the profession | Regulators judge whether firms had adequate safeguards | Document policies, approvals and vendor due diligence |
| Vendors | Publicly backed providers and programmes | Liability stays with the firm, not the attacker | Review contracts for security standards, breach notice and audit rights |
| Staff capability | Practitioners expected to use tools routinely | Human error and weak access controls remain common causes | Targeted training on confidentiality, access and phishing |
| Enforcement | Regulatory expectations rising alongside adoption | Fines upheld as proportionate and deterrent | Treat breach readiness as a board-level issue |
The Regulatory Gap Beneath the Strategy
The third piece of the picture comes from Parliament. The House of Commons Justice Committee has warned that the current regulatory framework fails to prevent serious consumer harm and has called for fundamental statutory reform. It also raised concern that refocusing the Legal Services Board on its core priorities could leave gaps in leadership on legal technology innovation and research.
That concern matters for the LawtechUK agenda. If the body responsible for overseeing the regulatory framework is pulled back to core statutory duties, and innovation leadership is left to government programmes and private providers, then the rules governing AI use in legal services may lag behind the tools themselves. Firms could find themselves adopting technology faster than the standards that should govern it are being written.
Why Consumer Harm and Data Security Are Linked
The Justice Committee's focus on consumer harm and the ICO's penalty against DPP Law address the same underlying problem from different directions. Clients rely on solicitors to protect information and to deliver competent advice. When either fails, the harm falls on the client first. A regulatory framework that does not adequately address technology risk, combined with rapid AI adoption, creates precisely the conditions in which harm can scale.
The current regulatory framework fails to prevent serious consumer harm, and fundamental statutory reform is required.
The Committee's warning, as reported by Legal Futures, is a useful lens for firms preparing for the next phase of change. Waiting for statute to define acceptable AI use is unlikely to be a viable strategy. Firms will need to build their own standards now.
Practical Steps for Firms in 2026
For managing partners, COLPs and heads of IT, the combined picture suggests a clear sequence of priorities:
- Audit before you adopt. Identify every system, including AI tools, that stores or processes client data, and confirm where that data sits and who can access it.
- Set a written AI use policy. Define permitted uses, prohibited inputs, approval routes and human review requirements, and make the policy known to all fee earners.
- Tighten vendor due diligence. Ask for evidence of security certification, data processing terms, incident notification timelines and audit rights before signing contracts.
- Test your breach response. A plan that has never been rehearsed will fail under pressure. Run tabletop exercises that cover data exposure, regulator notification and client communication.
- Train for the real threats. Phishing, credential theft and misdirected files remain common routes into firms. Training should be practical and refreshed regularly.
- Keep records of decisions. If a regulator asks why a control was or was not in place, documented reasoning is far more persuasive than memory.
Smaller firms face the sharpest version of this challenge. They may lack dedicated security teams, yet they are the audience LawtechUK is most likely to reach. Shared resources, reputable managed-service providers and clear guidance from professional bodies will matter for them.
Looking Ahead
The next phase of LawtechUK will be judged on whether it makes AI genuinely useful and trusted in everyday legal work. The DPP Law ruling will be remembered as an early marker of how seriously the courts and regulators take client data, whatever tools a firm uses to handle it. The Justice Committee's warnings suggest that the wider regulatory architecture is still being debated and may take time to catch up.
The most successful firms over the next two years will probably be those that treat trust as the foundation of technology strategy. Adoption without security invites penalties and reputational damage. Security without adoption leaves firms behind a market that is moving quickly. The task for UK law in 2026 is to build both at the same time, and to do so before the next breach makes the lesson for them.
