3 min read

FL #90: Is agentic AI the future?

Is the future autonomous systems that plan, reason, and act. From virtual project managers to due diligence agents, it promises efficiency but faces oversight challenges. Plus funding roundup and book recommendation.
FL #90: Is agentic AI the future?

✈️
Heading to ILTACON next week? I will be there too and would love to meet in person. Reply to this email, and let’s find a time to chat in Nashville.

The legal profession is no stranger to technological advancements by now, with generative AI already making waves in document review, contract analysis, and legal research. Is the future of AI-powered legaltech agentic AI?

This approach addresses some limitations of current AI systems and promises to transform the role of knowledge workers in law firms.

What is Agentic AI?

Agentic AI refers to systems that can act autonomously to achieve specific goals. Unlike traditional (or passive) AI models that respond to prompts or provide suggestions, agentic systems can plan, reason, and, most importantly, take actions to pursue objectives. This capability allows them to interact more dynamically with their environment and users, potentially redefining the landscape of legal tech.

Potential Applications in Law Firms

In the context of a law firm, agentic AI might manifest as:

  • AI Virtual Legal Project Manager: This agent could autonomously manage entire matters, ensuring appropriate scoping and budgeting and removing friction from work management. It could allocate tasks, track billable hours, manage resources, and suggest strategic adjustments to project plans based on progress and emerging issues.
  • Legal Research Agents: Instead of providing relevant case law, these agents could synthesize information, draft memos, and update case strategies based on new precedents or legislative changes.
  • Due Diligence Agents: These could independently review contracts, flag issues, and negotiate standard terms within predefined parameters.

The key distinction here is that these agents would not just suggest actions or provide information - they would take concrete steps to complete tasks involving human oversight only when necessary or desired.

Challenges

As great as all this sounds, there will be many issues to overcome in practice. The risk of significant errors without proper oversight cannot be overstated. An AI agent making an autonomous decision about case strategy or contract negotiation could have far-reaching implications if it misinterprets nuances or fails to consider all relevant factors (not to mention the lack of sufficient data to make the decision).

Given these risks, it's likely that the development of AI agents will occur in phases:

  • co-pilots (notice the small 'c'): the agent performs complex tasks but requires human approval before taking action. For example, a legal research co-pilot might draft a memo but wait for an attorney's review and approval before sending it to a client. This approach allows for the efficiency gains of AI while maintaining the critical oversight of experienced legal professionals.
  • Autopilot: As these systems prove their reliability and firms become comfortable with AI-driven processes, we might see the emergence of more autonomous agents. However, even in this future scenario, there must be oversight, particularly for high-stakes or novel issues.

Of course, agentic AI is still fringe technology while firms, regulators, and technologists consider its implications. Most successful near-term outcomes will require balancing automation and human expertise. AI can enhance efficiency and handle routine tasks, but human judgment and problem-solving skills remain invaluable.


πŸ€‘ Funding

July 2024 saw an unprecedented surge of mega-rounds for legal tech, with over $1 billion raised. Here are the biggest five rounds.

  • Clio's massive $900M Series F led by New Enterprise Associates (NEA) is now the largest legal tech investment of all time, valuing Clio at $3B.
  • Harvey, also raised a huge $100M round in July 2024 led by OpenAI.
  • Leya, an AI-powered legal workspace raised $25M in Series A, led by Redpoint Ventures.
  • netLex, a Brazilian CLM provider raised $24M focusing on AI integration and international growth.
  • LegalFly secured €15M in Series A. LegalFly, based out of Ghent, is an AI-native workspace to automate legal services.

Interesting look by Noah Waisberg comparing the Clio and Harvey raise.


Book recommendation

I’ve been reading β€œThe Legal MBA” by James Markham and Darren Mee and recommend it as an excellent reference.

Back in 2010 I read the Personal MBA - which provides learnings from an MBA in a single volume. Similarly, the Legal MBA aims to provide key concepts from a traditional MBA curriculum and apply them to legal services.

Legal professionals need to understand business concepts and the uniqueness of law firms, and the authors have done a great job of providing an easy-to-read reference. Whether your focus is project management, pricing, leadership, or, frankly, just understanding the business of a law firm, this is a great place to start.


What happens when you are in the middle of a revolution, but nothing radical appears to be happening? (see first link below)

πŸ”— In other news

Become a Fringe Legal member

Sign in or become a Fringe Legal member to read and leave comments.
Just enter your email below to get a log in link.