Orchestrating Intelligence: AI Agents in the Legal Space – Session Summary

Speakers: Lisa Erickson (Aderant), Matt Zerweck (Harvey), Adam Ryan (Litera), Joel Hron (Thomson Reuters) AI agents are goal-oriented systems designed to understand context, plan actions, and execute tasks autonomously. They function as sophisticated digital colleagues, capable of understanding objectives, utilizing available tools, planning execution, and seeking guidance when necessary. The key distinction from traditional AI is that users define the desired outcome, rather than dictating specific steps. These agents are proving to be transformative in the legal sector, enabling professionals to significantly increase their output and quality of work. They excel at amplifying the most human and challenging aspects of legal practice. As these agents gain autonomy, human oversight remains crucial. Future software will likely focus on optimizing the speed of verification, supported by transparent citation and source tracking. Current applications of AI agents in law include proactive email processing, responding to inquiries, and generating pitch materials. In document drafting, they are reportedly saving 50-70% of the time to reach initial drafts, while also ensuring greater consistency by incorporating firm and client preferences. Legal research, particularly deep dives, is described as a profound use case, demonstrating over 60% time savings and uncovering new arguments across different jurisdictions. Contract analysis is another key area, with agents identifying standard terms, flagging non-standard provisions, and proactively assessing risk within contract portfolios. The development of AI agents relies on a three-pillar approach:

• Planning and reasoning capabilities: The core logical processes that enable agents to strategize.
• Purpose-built tools: APIs specifically designed for agent interaction.
• Context provisioning: Access to relevant data, encompassing both proprietary firm data and third-party information.
Looking ahead, the ecosystem of AI agents is predicted to grow, with agents developing the ability to communicate and collaborate with each other more effectively. Visionary forecasts include agents proactively reaching out with suggestions based on incoming information. A critical enabler for this future is the availability of well-structured firm experience data. Successful implementation involves providing agents with robust information and context. It is advisable to start with simpler tasks before moving to complex workflows. Human review of agent output remains essential, and proper access controls are necessary to ensure data security. The effectiveness of agents is task-dependent, with some tasks requiring minimal guidance while others demand frequent support. The overarching takeaway is that the impact of AI agents in law will be even greater than currently anticipated, with the strategic transformation unfolding gradually as firms build foundational data and verification processes.

---

Actionable AI Strategy & Policy – Session Summary

Speakers: Sean Monahan (Harbor, Moderator), Sukesh Kamra (Torys LLP), Christian Lang (Lega), Anna Corbett (Akin Gump) Polling indicated that nearly half of the attendees are currently in the "piloting tools" stage of their AI adoption journey, actively testing but not yet fully deployed. Key debates and findings from the session highlighted a balanced approach to strategy and experimentation as the most effective path. Anna Corbett advocated for balancing governance with flexibility, while Sukesh Kamra emphasized the need for a clear strategy and objective, likening it to needing a map. Christian Lang, however, suggested that strategy could emerge through experimentation, stating, "We have absolutely no idea where this is going." Regarding policy development, Sukesh Kamra stressed the importance of establishing policy upfront in a regulated industry. Christian Lang focused on structural safety over written policies, and Anna Corbett proposed that policies should mature over time based on actual use. In terms of technology investment, Anna Corbett recommended investing in foundational AI tools that offer immediate productivity enhancements. Sukesh Kamra advised conducting a readiness assessment before making major investments, while Christian Lang suggested focusing on R&D and experimentation rather than large upfront investments. Views on transformational versus incremental change were mixed. Christian Lang posited that technical legal skills might become less relevant within 18 months, while Anna Corbett advised planning for both short-term efficiencies and long-term transformation. Sukesh Kamra noted that the approach depends heavily on leadership and organizational culture. Lightning round insights revealed that "politics" is considered harder to navigate than technology. The majority of attendees were opposed to the idea of a Chief AI Officer by 2026, viewing AI as an evolution rather than a role requiring a new C-suite position. Views on AI ownership structure were divided, with a cross-functional approach being favored. A three-step implementation framework was presented:
• Conduct Readiness Assessment: Evaluate change tolerance, structural readiness, define success metrics, and assess architecture, risk management, and training capabilities.
• Deploy Foundational AI Tools: Begin with productivity-focused platforms, help lawyers understand AI basics, and balance experimentation with platform strategy.
• Implement Flexible Governance: Establish baseline safety requirements, avoid letting risk concerns block opportunities, and create structural safety rather than relying solely on policy compliance.
Notable quotes included: "The fundamental barrier to adoption [is] getting lawyers to use prompts." Regarding change drivers, Christian Lang stated, “Who gets it and who uses this are going to drive the change more than any one position.” On transformation, he added, “Anyone who truly believes this is incremental improvement technology and plans to do business the same way in five to ten years is going to be out of a job.”

---

Fun Run with Ari Kaplan

As part of ILTACon Day Three, Draftwise organized a fun run featuring legal tech expert and consultant Ari Kaplan, adding an energetic and community-building element to the event.

---

Thanks to the Draftwise team for their coverage of ILTACon.

---

Originally published at Artificial Lawyer on 14 August 2025.

---

Frequently Asked Questions (FAQ)

What are AI agents in the legal context? AI agents are sophisticated, goal-oriented systems that understand context, plan actions, and execute tasks autonomously. They act like a "good co-worker" by understanding tasks, available tools, and planning execution, seeking guidance when needed. The key difference is that users specify what they want to achieve, not the specific steps to take. How do AI agents benefit legal professionals? AI agents can significantly amplify the capabilities of legal professionals, allowing them to accomplish more with higher quality. They handle complex and time-consuming tasks, freeing up human professionals to focus on more strategic and nuanced aspects of their work. What are some current use cases for AI agents in law? Current applications include email processing (responding to inquiries, generating pitch materials), document drafting (saving 50-70% of time to initial drafts), legal research (achieving over 60% time savings and discovering new arguments), and contract analysis (identifying standard terms, flagging risks). What is the role of human oversight with AI agents? Human oversight is critical, especially as agents become more autonomous. Verification processes are essential, and future software will likely focus on optimizing this verification speed, ensuring transparency through citation and source tracking. What are the key considerations for implementing AI agents in a law firm? Key considerations include providing agents with sufficient context and information, starting with simpler tasks, ensuring human review of agent work, implementing proper access controls, and understanding that results vary based on the task complexity and guidance provided. What is the future outlook for AI agents in the legal space? The future anticipates ecosystems of collaborating AI agents, proactive intelligence where agents reach out with suggestions, and a greater reliance on structured data sets of firm experience for successful agent performance. How does a firm approach the strategy for AI adoption? A balanced approach between strategy and experimentation is often recommended. While a clear initial strategy is beneficial, allowing room for emergent strategy through pilot programs and learning from actual use is also crucial. What are the essential steps for implementing AI in a legal practice? A recommended framework includes: conducting a readiness assessment, deploying foundational AI tools to boost productivity, and implementing flexible governance structures.

Crypto Market AI's Take

The advancements in AI agents, as discussed in the legal sector, mirror the transformative potential seen across various industries, including finance and cryptocurrency. Our platform, Crypto Market AI, leverages similar AI-driven principles to provide sophisticated trading bots and insightful market analysis. The emphasis on providing agents with robust data and context for optimal performance aligns directly with our approach to building intelligent trading tools. As legal professionals aim to amplify human potential with AI, we focus on empowering crypto traders with data-driven insights and automated strategies to navigate the dynamic digital asset markets. Understanding the core mechanics of AI agent implementation, such as the three-pillar approach to training and the importance of structured data, is vital for developing effective AI solutions, whether in law or finance.

More to Read:

• AI Agents: The Future of Business Automation
• Navigating the AI Revolution in Finance
• Understanding Cryptocurrency Trading Bots

Originally published at Artificial Lawyer on 14 August 2025.