In the world of law, the advent of generative AI was expected to revolutionize how information is processed and managed. Traditionally, the search, retrieval, and evaluation of information within law firms were meticulous, manual tasks. Lawyers had to sift through various versions of documents before proceeding with cases. However, with AI, a much quicker, automated process delivers polished answers, often without revealing the chaotic state of the underlying data.

Kurt Thies, global head of Information Governance Strategy at legal services consultancy Morae, highlights a significant issue: “We’ve always known that organizations don't have confidence in their information. But the consequence of that now has hit a whole new level.” According to research by Morae, a mere 26% of legal leaders are confident that their organization's information is primed for effective AI analysis, and only 37% are aware of where all their legal information is stored.

This gap underscores the critical need for robust information governance. Before integrating more AI tools, firms must address the prevalent issue of 'ROT' — redundant, obsolete, and trivial data. When AI systems access such disorganized data, they can produce outputs based on outdated or irrelevant information, such as multiple versions of a single contract or outdated agreements.

Moreover, the risk extends beyond inefficiency. Poor information hygiene can expose firms to significant risks. Only 30% of legal leaders are confident that their information is correctly classified as confidential or sensitive. AI struggles to differentiate, potentially leading to severe data breaches. For instance, Thies describes a scenario where HR files on a shared drive could be inadvertently exposed by an AI system if not properly indexed or classified.

Proper information governance is more than a compliance requirement; it is essential for AI effectiveness. “AI lacks the organizational context unless it is given parameters through the governance of that information,” Thies explains. Without a clear retention policy that is actively enforced, irrelevant records remain accessible to AI, muddling the system’s outputs with outdated or trivial data.

Despite significant investments in AI, only 33% of senior legal professionals trust the results of AI-assisted legal work. Many are concerned that the costs of human verification might negate the productivity benefits that AI promises. Thies stresses that information governance should not be seen as a barrier to innovation but as a prerequisite for reliable AI outcomes.

The journey towards trustworthy AI in legal practices demands a candid evaluation and restructuring of how information is maintained. “The industry can get there, but only by confronting the condition of its information first,” Thies concludes. As it stands, the path to effective AI integration is clear, yet many firms remain unprepared to take the necessary steps.