Linklaters hosted a report launch on 2 December, organised by She Breaks the Law (SBTL) and Next Hundred Years, aimed at amplifying women's voices in the rapidly evolving landscape of Generative AI (Gen AI) within the legal sector. Aedamar Comiskey, senior partner at Linklaters, inaugurated the evening by highlighting findings from the report which revealed a widespread lack of confidence among women in adopting Gen AI technologies, potentially leading to a knowledge gap that could hinder progress.

The SBTL, a global network focused on supporting women professionals in law, was represented at the event by Chief Marketing Officer Helen Burness and director Lucie Allen, who both emphasised the pressing need to ensure women's perspectives are integrated into the development and application of AI technologies. They expressed concerns regarding the current underrepresentation of women in positions of influence within AI development.

A central theme discussed was the inherent biases present in large language models (LLMs). A notable example shared by Shilpa Bhandarkar, Linklaters’ partner and Gen AI chair, involved a female lawyer who struggled to adjust the tone of a professional email until she instructed the AI to "write like a man." Such instances raise critical questions regarding the representational data upon which these models are trained. Fiona Phillips, head of legal – digital information and data at UBS, recounted her experience leading an AI Act training session for engineers, where an exercise to depict a "fun" lawyer yielded only male representations, and when she requested a female portrayal, the results were starkly sexualised.

Christina Blacklaws, former president of the Law Society of England and Wales and chair of the legal tech startup organisation LawTechUK, added that only 14% of scale-ups and startups are female-founded, underscoring the importance of diversity in achieving varied outcomes within the industry.

Dana Denis-Smith, founder of Next 100 Years and provider of flexible legal work through Obelisk Support, encouraged women to engage actively with AI models like ChatGPT in order to address biases. She shared an exercise where her daughter queried ChatGPT about the world's greatest leaders, resulting solely in male names. Denis-Smith remarked on the necessity of ongoing engagement in AI training to help shape its outputs, stating, "If we don’t call it out, it won’t learn."

Contributing to the discussion, Linklaters partner Marc Harvey, who described himself as a 'male ally,' noted that the firm's Gen AI steering committee comprises predominantly women. He also shared that the bulk of the firm’s chatbot usage comes from female users, emphasising the significance of ensuring equitable access to technological advancements. Harvey advocated for firms to systematically integrate new technologies into daily practice and institutional processes.

The panel collectively voiced the need to draw lessons from the past as the legal sector navigates the opportunities and challenges posed by Gen AI. Bhandarkar stressed that learning from historical oversights is vital to avoiding repetition of the same mistakes, while also calling for collaborative policies from both governmental authorities and tech companies. The need for proactive measures from organisations like The Law Society and the City of London Law Society in this domain was highlighted as an essential step in facilitating a more inclusive AI landscape in the legal profession.

Source: Noah Wire Services