With everyone talking about ChatGPT and the implications of AI tools for the future of various professions, now is an opportune time to consider how AI might change legal recruiting. We at Lateral Link have been actively engaging with this question for years: in fact, we have a sister company called Haistack.AI that is developing AI products for the legal recruiting industry.
So for the latest episode of the Movers, Shakers & Rainmakers podcast, we invited Haistack.AI Chief Technology Officer Michael Heise to discuss the possibilities and limitations of AI for law firms and legal recruiters. Mike educated us on the likely implications of AI for our industry and described the logic behind the product that Haistack.AI is currently building.
Mike is a seasoned legal tech innovator with a deep understanding of Biglaw firms. Prior to joining Haistack.AI, Mike held software leadership roles at Cooley and Covington & Burling. As he explained on the podcast, he is married to an attorney, and it was his wife who first sparked his interest in legal sector innovation.
AI can be a valuable tool
Mike explained that AI has the potential to assist lawyers with a broad range of tasks. For example, a litigator could rely on AI tools to set out the basic structure of a brief, allowing the lawyer to dedicate more of her time to the higher-value tasks of refining arguments and tailoring them to be most persuasive based on the unique facts of the case. As Mike puts it: “AI is not going to replace you. The person who knows how to effectively use AI is going to replace you.” AI tools will become increasingly sophisticated, but human judgment will remain essential for crafting the strongest and most original arguments.
Similarly, AI is well suited to help recruiters — both within firms and outside them — to more efficiently identify high-potential candidates. By reducing the time a recruiter spends on manually trawling through candidate profiles, AI can enable the recruiter to gain a deeper understanding of the high-potential candidate pool and the relative strengths of the candidates within that pool.
The Haistack.AI vision
As an example of how AI promises to make recruiting more targeted and efficient, Mike described the product that the team at Haistack.AI is building. It entails creating three essential models: (1) profiles of lawyers currently working at the firm that is using Haistack.AI in its hiring process; (2) profiles of lawyers working outside that firm; and (3) profiles corresponding to the specific roles for which the firm is recruiting. By comparing the profiles of lawyers previously hired by the relevant practice group and office with the profiles of external lawyers, the algorithm can instantly generate a list of high-potential candidates and an explanation for why those candidates appear to be a good fit. Moreover, the AI will use Lateral Link data to screen out candidates whom the firm has previously considered and determined not to be a fit. Finally, the tool will give some indication of the extent to which the leading candidates are likely to be in demand at other firms seeking to fill similar vacancies, alerting the hiring firm to the need to move quickly where a candidate is likely to be in especially high demand.
With the assistance of the Haistack.AI tool, the recruiter managing the search will immediately see how the algorithm matched a candidate’s qualifications and experience to those of current members of the group. This is where human judgment comes in. The AI accelerates the first step of identifying a shortlist, but the law firm’s recruiting and attorney professionals must assess whether the shortlist fits their needs, through interviews and other more traditional evaluations.
Mike noted that in addition to generating lists of promising candidates, the Haistack.AI tool could also help identify current members of a firm who are in especially high demand relative to what the broader lateral market is seeking. In alerting a firm to attorneys who are at greater risk of leaving, the tool can help nudge a practice group to be more proactive about taking steps to keep valuable team members happy.
AI is not a panacea
Mike also explained the importance of recognizing the limitations of AI and of not buying into the excessive hype that frequently surrounds promising technologies in their early stages. AI will not solve all hiring problems. To take just one example, the inputs for AI models like the ones that Haistack.AI are building are composed of historical data — the models are designed to replicate the firm’s past hiring decisions. To the extent the past hiring was suboptimal, such as through failing to hire qualified diverse candidates, the AI tool will not correct the problem. Instead, it is important for the human users to be thoughtful about patterns in past hiring that they do not wish to replicate and make an active effort to change them.