Corporate legal departments are moving beyond experimenting with generative AI and into a phase where AI systems are operating inside real legal workflows. PYMNTS reports that AI agents are already reviewing contracts, monitoring regulatory changes and performing first-pass discovery at law firms and corporate legal teams. This changes the general counsel's role: the question is no longer simply whether AI is useful or safe enough to try, but how to govern systems that are already running and can make decisions with limited human intervention.
That shift creates new forms of legal and operational liability. Tools such as Microsoft's Legal Agent can analyze documents, draft edits and review contracts with significant autonomy. While that can increase efficiency, an autonomous system can also make a consequential mistake faster than a human oversight process can catch it. PYMNTS highlights the growing use of AI addenda in vendor contracts, covering issues such as violations of law, acts or omissions by AI systems and liability when an AI agent causes harm. Contracting around AI behavior is therefore becoming an important responsibility for general counsel.
The legal-technology market is also moving toward domain-specific AI agents. Harvey, for example, has introduced tools that allow law firms to create customized agents for particular legal practices. According to the article, Harvey serves more than 700 customers across 58 countries, with applications spanning contract analysis, due diligence, compliance monitoring and litigation support. The strategy reflects a broader industry trend: rather than relying exclusively on general-purpose models, enterprises are increasingly looking for AI systems that understand the terminology, processes and requirements of a particular profession.
The adoption figures show how quickly this transition is happening. A report from FTI Consulting and Relativity found that 87% of corporate legal departments reported using generative AI in 2026, compared with 44% in 2025, while the proportion with a formal technology roadmap increased from 25% to 53%. Summarization, contract-clause identification, e-discovery, document review and contract drafting are among the leading uses. The broader message is that legal departments are becoming AI governance centers, not just AI users: they must establish permissions, accountability, auditability, vendor protections and human oversight as AI moves from an experimental tool to an increasingly autonomous member of the legal workflow.
