Mid-market AI adoption trends (sponsored content)
Small and medium sized law firms are at a pivotal moment in the adoption of generative AI, yet most remain cautious, even as client expectations surge. Surveys reveals that while 26% of legal organisations are actively using generative AI - up from 14% in 2024 - actual integration into core legal workflows remains patchy and piecemeal.
In this article, Akber Datoo, CEO and founder of D2 Legal Technology (as well as the co-chair of the Law Society’s Technology and Law Committee), shares first-hand insights to the sector’s current state of AI adoption, level of understanding, use cases, barriers, next steps and critical considerations that will support firms to embrace AI effectively, ethically and with confidence to optimise their businesses.
Current State: Limited Adoption, Patchy Implementation
Law firms across the legal sector are confronting the realities of AI. Not just the technology, but the strategic leadership decisions that come with it.
While some law firms have followed a strict no-usage approach on AI, most have limited its usage to personal productivity tools (like Microsoft Copilot), yet without clear boundaries over what tasks require AI use and what tasks do not. Often ChatGPT is not permitted for business use – without addressing the blurred line between personal use and learning the basis of an area related to a client matter (often interpreted as simply – don’t put client documents into ChatGPT). Even where firms have enabled AI functionality in their core practice or document management systems, uptake is modest and governance frameworks are largely absent. As a result, these tools have yet to move beyond experimentation and firms have yet to unlock their full potential.
A 2024 Thomson Reuters survey supports this narrative: 63% of lawyers have tried generative AI, mainly for summarising and drafting, but only 15% have integrated it into daily workflows. Similarly, the ABA TechReport 2024 describes AI usage as “relatively limited to certain functions.”
Understanding: A Knowledge Gap Among Partners
Awareness of AI’s potential varies widely. Some partners see the time-saving and analytical power of large language models (LLMs), but for many, understanding of key concepts – model types, prompt engineering, hallucination risk, and distinctions between open-source versus fine-tuned systems – is lacking. In fact, in many cases, there is a lack of understanding of the broad shift of technology that is AI – from old fashioned AI, to predictive AI, to generative AI.
This knowledge gap is compounded by the absence of clear AI policies, leaving firms at a disadvantage. Without formal guidance, employees remain uncertain about which AI systems they can use and for what purposes. A particularly concerning dynamic has emerged: when employees do experiment with AI tools and achieve time savings, they often fail to communicate this to management. In some cases, because AI usage isn't formally encouraged, employees may even report longer completion times to maintain billing and management expectations.
Use Cases: Routine Tasks, Document Automation, Knowledge Management
Firms are currently exploring low-risk use cases for AI, including:
• Automating administrative tasks (e.g., email triage, time entry and file opening)
• Document summarisation and clause extraction
• First-draft contract generation using template pre-population
• Enhancing knowledge management with client-specific dashboards
While these applications promise efficiency gains, firms need to consider which existing processes can be streamlined and optimised before applying AI.
Barriers: Technology, Culture, and Skills
Several hurdles block AI adoption:
• Legacy system integration challenges
• Cultural resistance to change
• Concerns about cost and unclear ROI
• Confidentiality, privacy, and regulatory risk
• Scarcity of in-house technical expertise
• Lack of structured training and limited time to experiment
The cultural challenge is particularly acute: busy fee-earners understandably struggle to find incentives and the time to adopt new tools, especially when it’s unclear whether time spent on AI-led matters will count towards billing targets and firm profitability.
No comments yet