On the theme of “The Intersection of Law & Technology: Turning Innovation into Advantage”, LegalTechTalk held a panel discussion titled “How to Make Smart Tech Strategy Choices in 2026 and Beyond” to explore how most in-house legal teams have accumulated their technology stack reactively, one tool at a time, in response to individual pain points rather than a coherent strategy. The result is a patchwork of platforms that do not talk to each other, create duplication of effort, and leave lawyers doing manual workarounds. This session moved beyond the audit to ask a harder question: not just whether your stack fits today, but whether it is built for what is coming next.
The question of legal AI is no longer whether it has value, but where to place the smartest bets next. This session cut through the noise to examine where genuine value is being created versus merely claimed, and what needs to change now to scale with confidence rather than catch up under pressure.
Moderated by Oliver Attinger, Editor & Founder, Non-Billable, the panel consisted of Thibaut Gregoire, Executive Vice President- General Counsel, Strategic Initiatives, M&A, Marketing & Communications, Legal Operations, Mastercard; Joanna Murphy, Founder, Rethink Legal; Rhys Hodkinson, Chief Revenue Officer, Definely; and Hélder Santos, Global Head of Legal Tech & Innovation, Bird & Bird.
Kickstarting the session, Oliver Attinger introduced the panellists and outlined that the session was premised on the fact that most legal teams might not have built their tech stacks strategically; they had accumulated one tool at a time.
Building, Buying, and Pondering: Thibaut Gregoire
Answering the question about legal tech at Mastercard, Thibaut explained that Mastercard’s legal technology strategy begins not with software procurement but with a broader transformation plan of the legal team. The starting point was understanding how the legal team should operate in the future, identifying use cases, assessing existing systems, and determining where genuine gaps exist.
He underscored that their strategy had been: “Building, Buying, and Pondering”. He explained that they had all the basic tools, not just for the legal team, but rather for the entire company. There was an organization-wide active conversation about the tools being used by all departments and how else the existing ones could be utilized without adding additional tools into the mix.
Selecting the right tech: Joanna Murphy
Regarding what mistakes buyers make while buying tech, Joanna remarked that though buyers had a list of requirements, the list was mostly a features list. For example, they mention features like generating documents based on a template, or comparing documents with model documents, etc.
Drawing on a contract lifecycle management project she had advised, Joanna explained how detailed questioning about document automation immediately eliminated several vendors that initially appeared suitable. Such specificity, she argued, was essential for meaningful comparison.
She also highlighted shortcomings in pilot programmes. In many organisations, technology testing is conducted informally alongside lawyers’ existing workloads, resulting in inconsistent evaluations and subjective assessments. To address this issue, she suggested a structured testing methodology involving clearly defined success criteria, feature or functionality being examined, standardised test data, and participation from multiple users assessing the same functions.
Building trust with legal buyers: Hélder Santos
Speaking from the perspective of a global law firm, Hélder outlined the factors that influence technology purchasing decisions. Security and trust, he explained, remain fundamental prerequisites. However, beyond technical assurances, trust is strengthened when vendors demonstrate transparency regarding product limitations and development roadmaps.
He criticised what he described as “demo theatre,” wherein a new tech product performs flawlessly during demonstrations but encounters difficulties when integrated into real-world legal workflows. Thus, trust was built with vendors when they addressed the arising day-to-day problems.
According to him, successful vendor relationships are built on transparency, mutual understanding, and a willingness to address challenges collaboratively.
Legal technology fatigue and growing complexity: Rhys Hodkinson
Continuing the discussion, Rhys drew attention to a growing challenge facing legal organizations, i.e., technology fatigue. While the market continues to generate innovative solutions and new vendors enter the market daily, he observed that legal professionals are increasingly overwhelmed by the sheer volume of products competing for attention. Lawyers are expected to not only do their day jobs but also evaluate new technologies and adapt continuously as vendors release frequent updates and enhancements.
He also highlighted how AI had, in some cases, increased workloads rather than reducing them. Historically, legal work was constrained by human capacity. Generative AI has dramatically expanded the volume of work that can be produced, but lawyers remain responsible for reviewing and taking accountability for every output.
“A lot of firms and organisations are starting to do a much deeper use case dive, a much more understanding of saying, these tools are fantastic, and we’ve seen real gains, but they’re also bringing up new issues or new things that we didn’t previously think we’re going to encounter.”
The industry, he noted, is beginning to move beyond excitement about AI capabilities toward more mature discussions about practical implementation mechanisms and challenges.
What tools to use: Thibaut Gregoire
Answering a question about small legal tech vendors, Thibaut underscored that while there were many such vendors pitching their products, some critical aspects for his company were security, privacy, and governance. He emphasized responsible use of every tool and driving user adoption.
Thus, he explained that he was striving for the right balance between short-term quick wins and developing momentum in the AI journey and the long-term strategy of building qualitative data legs, which will aid access to evolving technology. He illustrated that Mastercard’s use cases of AI were regarding due diligence, contract or document drafting, redlining, and legal research.
Role of technology: Hélder Santos
Addressing broader questions about the role of technology in legal services, Hélder explained that though tech had always been a part of their work, using the new tools was the correct method as they offered many great features. He added that clients wanted their issues addressed with the best legal services, augmented by technology, and providing value.
He emphasized that technology is complementary to professional expertise rather than a substitute for it. The firm’s objective was not to maximise the number of tools lawyers use but to help professionals deliver greater value to clients. This requires substantial investment in education and training focused on how technology can enhance legal services rather than merely automate tasks.
Regarding adoption, he underscored that the issue did not lie with tech, but rather with usage as people attempted to integrate new tech with old processes and mechanisms.
Hélder rejected simplistic approaches to measuring AI adoption through metrics such as prompt counts or usage targets. Lawyers should be recognised for delivering value, solving complex problems, and contributing to organisational learning, not for generating large numbers of AI interactions.
Adoption and usage
Joanna agreed that adoption remains one of the most significant challenges facing legal organisations. However, she challenged the notion that adoption problems stem solely from user resistance. Instead, she characterised many adoption failures as design failures. Where technologies fail to address meaningful pain points or introduce unnecessary workflow friction, users naturally disengage. Effective adoption, therefore, requires careful consideration of workflow integration, usability, and the practical realities of legal work.
She explained that most of her clients, which were typically smaller or medium-sized organizations, adopted and increasingly showed interest in general-purpose tools for brainstorming, summarizing, etc. She characterised such use as low-hanging fruit. As clients were still not adopting AI for tasks like contract review, they did not have a protocol for usage, an understanding of good outputs, and a consistent approach to negotiation. Thus, without ground rules, such enterprises could not give guardrails for AI use to get the right results.
Speaking from his experience at Mastercard, Thibaut explained that Mastercard’s legal team had become one of the organisation’s leading adopters of AI technologies through extensive training programmes, identifying champions, and a sustained focus on user engagement. For him, adoption depends on helping users understand what value technology delivers in their daily work. He underscored the importance of building strong foundations, including workflow standardization, data governance, document organization, using correct tools, decision-making playbooks, and enterprise-wide collaboration.
Governance and Risk of AI
Regarding AI governance, Hélder argued that governance should not be viewed as a policy document stored in a folder. Instead, governance must be evaluated by how lawyers behave under pressure, particularly when deadlines are approaching, and shortcuts become tempting.
At Bird & Bird, governance is supported through extensive training initiatives, experimentation environments, and carefully controlled access to approved tools. Emphasizing the need for flexibility, he mentioned that the firm established innovation sandboxes where legal professionals can explore new technologies within defined boundaries. He opined that while governance frameworks establish boundaries, organisations must also create space for innovation and learning.
From the perspective of Mastercard, Thibaut explained that the company has taken a proactive position in supporting AI use, provided that robust governance, security, and accountability measures are in place. He described relationships with tech vendors as partnerships in which both sides collaborate to evaluate emerging technologies, share lessons learned, and identify efficiency gains.
While governance frameworks remain essential, he noted that discussions about liability and accountability are not fundamentally different from those that have historically accompanied other forms of technology adoption.
Rhys added that the most important issue facing organisations today is operationalising governance rather than merely drafting policies. As the cost of generating legal content approaches zero, the true value increasingly lies in review, validation, and accountability. Lawyers continue to serve as the final safeguard against errors, inaccuracies, and unacceptable risks. He described this challenge as the separation between the “generation layer” and the “review layer.”
He reiterated that while AI can create content at unprecedented speed, professional responsibility remains firmly with the individuals and organisations that approve and rely upon those outputs. Consequently, legal teams must carefully consider how different technologies fit within broader governance architectures and ensure that outputs meet their criteria.
Concerns surrounding AI tools
Joanna observed that concerns about security, privacy, and trust continue to influence technology adoption decisions. Interestingly, she noted that the rise of large foundational AI models had, in some respects, made vendor due diligence more complicated.
She explained that where specialist legal technology providers previously exercised greater control over their technology stacks, organisations now increasingly depend on third-party models whose training methodologies and governance frameworks may be less transparent. As a result, many legal teams remain skeptical about how and where sensitive information is processed.
Thus, she suggested that organisations need stronger internal capabilities to assess AI outputs, establish guardrails, and develop confidence in responsible deployment.
Looking ahead
Thibaut predicted that AI capabilities will increasingly become embedded within business workflows rather than existing as separate applications. In the law department, there would be proactive legal support integrated directly into organisational processes. He added that there will also be a network of additional solutions but expressed concern that user experience will suffer as technology will become increasingly layered, thereby risking a reduction in adoption.
For Hélder, the future lies in reducing fragmentation and simplifying lawyers’ daily experiences. He predicted continued consolidation within the market and argued that lawyers should not be required to navigate multiple disconnected platforms merely to perform routine tasks.
Rhys echoed these views, suggesting that vendors must move beyond selling standalone products and instead demonstrate how their technologies integrate seamlessly, together, and into broader ecosystems. Future success, he argued, will depend on delivering functionality wherever lawyers choose to work rather than forcing them into predefined environments.
According to Joanna, she predicted increased adoption with the organizations realising the importance of laying foundations, governance, streamlining operations, and standardization.
Last piece of advice
Hélder encouraged innovation leaders to engage directly with users and understand the problems they are trying to solve before introducing new tools, changes, or processes.
“Innovation is not the objective, it’s the enabler of the transformation, and we shouldn’t be losing sight of the end objective.”
Reiterating the importance of data hygiene, Rhys emphasised the enduring importance of data quality, change management, strategizing, and user adoption.
Joanna concluded by cautioning organisations against treating technology as a simple add-on to existing ways of working. Instead, legal teams must rethink processes, build strong foundations, and align technology investments with clearly defined objectives.
