By Karen Waldron, LexisNexis.
The legal profession is built on trust, which creates a very specific challenge for AI. Even the most advanced technology will only matter if lawyers trust the answers, clients trust the process and firms can stand behind the work.
For CTOs at leading law firms, the question has moved from whether AI should be adopted to how AI can be made trustworthy across the business.
Our latest survey of private practice lawyers in the UK and Ireland shows how quickly this shift is happening. Four-fifths of lawyers at large firms now use AI for legal research, while roughly two-thirds use it for knowledge management, large-scale document review, document analysis and client document drafting.
This level of adoption is significant, but it does not necessarily indicate maturity. Only 30% of legal professionals at large firms said AI is embedded in their team’s strategy and operations.
This gap creates inconsistency. AI may be used across research, drafting and knowledge work without clear guidance on when it should be used, which tools are approved, how outputs should be reviewed, or where accountability sits.
As Alex Bazin, COO and CTO at Lewis Silkin, says: ‘Very few firms have rewired their underlying processes, feedback loops, and expectations to drive consistency of usage. Until that happens, the gains from AI will stay trapped with individuals rather than compounding across the firm.’
This is why the CTO’s role is becoming more strategic. The task is not simply to provide access to AI tools, or to encourage experimentation, but to create the conditions in which AI can be used consistently, safely and with confidence.
—
—
AI adoption is not the same as AI integration
As lawyers use AI for research, document review, analysis, drafting and knowledge management, the consequences of poor implementation become higher.
Hélder Santos, Head of Legal Tech and Innovation at Bird & Bird, says the challenge is connecting AI carefully to the way firms actually operate: ‘AI needs to be carefully connected to a firm’s systems, processes, and ways of working, in a way people actually trust and use.’
That connection is not always straightforward, as ownership of AI remains unclear inside many firms. It may sit with IT, innovation, risk, knowledge or individual practice groups, depending on the use case. The result is that tools can be bolted onto existing workflows rather than embedded as a seamless part of legal or administrative processes.
Much legal work is still bespoke and complex and does not lend itself to simple end-to-end automation. Integration must involve legal teams in identifying where AI genuinely improves speed, consistency or quality and where human judgement must remain central.
Jason King, Head of Platform at Taylor Rose, says poor data quality is one of the biggest barriers to meaningful integration: ‘Legal data is often fragmented across systems, inconsistently labelled, and embedded in siloed workflows, which limits what AI can reliably do at scale.’
This is where AI integration becomes closely linked to broader digital transformation. Firms cannot expect AI to deliver consistent value if the knowledge, documents and data it depends on are fragmented or difficult to access.
Lawyers will not trust any old AI
The main barrier to AI integration is trust. Despite the legal profession’s risk-averse reputation, lawyers are clearly willing to use AI where they see value. What they are less willing to do is rely on systems they cannot understand, verify or control.
Our survey found that 85% of legal professionals at large law firms are concerned about inaccurate or fabricated AI outputs.
For Oliver Bethell, Chief Technology Officer at Travers Smith, trustworthy AI has two dimensions: ‘Trusting that the data is secure and trusting that the model will produce reliable answers.’
Both dimensions are essential. Lawyers need confidence that client data is protected through enterprise-grade privacy, information security and retention controls. They also need confidence that the answer produced by the system is grounded in reliable material and can be reviewed properly.
This is also why legal-specific AI tools remain important. Across all respondents, 72% said they feel more confident using AI that is grounded in legal sources. Among lawyers at large firms, that rises to 79%, and among in-house corporate lawyers it rises to 85%.
Generic AI may be useful for lower-risk or more general tasks, but core legal work requires a higher standard. Lawyers need AI that is secure, grounded, auditable and designed around the realities of legal decision-making.
Trust is now a client issue
Clients are becoming increasingly interested in how firms use AI, what safeguards are in place, and how legal work is reviewed. For in-house counsel, the top strategic priority in 2026 is setting rules and safeguards for AI tools and products, with 39% stating this is a priority and 38% saying this is something they are already doing.
Nigel Lang, CIO at Fieldfisher, says transparency is central to building trust: ‘Lawyers need to understand how the AI arrived at a conclusion. Security, auditability and the extent to which humans remain in control all shape whether a system feels trustworthy.’
This means firms need clear answers to practical questions. Which AI tools are approved? What data can be used? Is client data used for model training? How are outputs checked? Who remains accountable? How is AI-assisted work recorded, reviewed or disclosed where necessary?
Introducing the Four-Layer Trust Stack
If trust is the central barrier to AI integration, firms need a practical way to build it. To help frame that challenge, we have developed the Four-Layer Trust Stack for legal AI implementation.
The first layer is infrastructure trust. This covers the security and privacy foundations of AI use, including architecture, access controls, risk mitigations, data retention and auditability. This is the layer CTOs are likely to be most familiar with, but it remains fundamental.
The second layer is technical trust. This relates to the quality and depth of the AI tool itself. What content is it grounded in? Are the sources authoritative? Can users access citations that add value? Where does the underlying content come from? Can the firm test, evaluate and monitor the quality of the outputs?
The third layer is workflow trust. This looks at where AI fits into the work lawyers actually do. It asks how AI supports research, drafting, review, document analysis or knowledge management and what steps are required before an output can be used safely. The key metric here is not simply speed. It is time-to-safe-answer: how quickly a lawyer can reach a result that is not just useful, but checked, understood and defensible.
The fourth layer is human trust. This includes change management, incentives, training, confidence, behaviour change and professional culture. Lawyers need to know how to use AI, when to use it, when not to use it and how to challenge what it produces.
Taken together, these four layers show why AI trust cannot be solved by technology alone. It depends on secure systems, authoritative content, embedded workflows and capable people. Remove any one of those layers and confidence starts to weaken.
Read the full report: In CTO we trust
–
About the author: Karen Waldron is a Senior Director for Product Development at LexisNexis UK.
—
[ This is a sponsored thought leadership article by LexisNexis for Artificial Lawyer. ]
Discover more from Artificial Lawyer
Subscribe to get the latest posts sent to your email.