On September 18, during WLG's regional conference in New York City, members from across the Americas gathered for a day and a half of programming focused on business opportunities and shared challenges. One of the most dynamic sessions featured Horace Wu, Founder & CEO of Syntheia, who led a discussion on AI Adoption in Law: Meeting Client Demands Responsibly.
Setting the Stage
Wu shared his journey from transactional lawyer to legal tech entrepreneur, explaining how a turning point in 2017—when researchers discovered new ways for machines to interpret text—set the foundation for today's generative AI tools. He likened AI adoption in law to the shift from steam engines to electric motors: AI is most valuable when inserted into the right parts of a workflow, not when swapped in wholesale to replace existing systems.
Current Adoption and Challenges
Law firms are already experimenting with AI tools, though satisfaction varies. Partners often praise platforms like Harvey or Microsoft Copilot, but associates—who must transform imperfect outputs into finished work—express greater skepticism. Wu suggested a practical rule of thumb: if a task is still valuable when it is only “60 percent right,” AI can play a meaningful role. If not, it risks adding more work than it saves.
Training the Next Generation
Delegates raised questions about the impact on associate training. Would younger lawyers who rely on AI lose the chance to build core analytical skills? Wu acknowledged the uncertainty, noting that AI has been observed to cause “smart people get smarter; dumb people get dumber.” Or, to put it more politely, less experienced lawyers risk getting worse if they don't know how to prompt correctly. He urged firms to think about layered training: building technical skills to use AI responsibly, applying the tools to specific workflows, and ultimately reshaping business models, pricing, and client relationships in response to the technology's growing role.
Practical Use Cases
Participants shared how they are already using AI in their practices. Some treat tools like Harvey as a junior associate, using it to draft agreements, compare legislation, or prepare conference bios. Others find value in translation, drafting first versions of client letters, or supporting internal training and feedback exercises. Many agreed that the greatest efficiencies currently come not from complex legal analysis but from the “business of law”—non-billable tasks where risks are lower and time savings are substantial.
Concerns and Client Expectations
Despite these opportunities, concerns remain. Confidentiality risks were top of mind, as was the potential erosion of critical thinking if AI becomes a crutch. Delegates also debated how clients will react once they know AI is part of their legal services. Will fees come under pressure? Will clients turn to AI directly and ask firms only to verify the outputs? These questions highlight the delicate balance between innovation and responsibility.
Looking Ahead
Wu closed by offering a timeline for the future. In the next two years, generative AI will become a feature in nearly every tool lawyers use, though its outputs will still demand careful review. Within five years, adoption will be widespread and firms failing to keep pace will fall behind. And ten years from now, client behavior may shift dramatically, with some bypassing law firms altogether for routine tasks. In that environment, Wu emphasized, the differentiator will remain the same: lawyers' expertise, judgment, and ability to guide clients responsibly through a fast-evolving technology landscape.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.
