From Copilot to Colleague: Trustworthy Agents for High-Stakes - Joel Hron, CTO Thomson Reuters

1.7K views · Jul 23, 2025 · 19:45 min · Watch on YouTube ↗

Takeaway

For high-stakes domains, agentic AI requires expert-driven evals, tunable agency dials, and exposing legacy systems as agent tools.

Summary

Thomson Reuters CTO: northstar has shifted from 'helpful' assistants to 'productive' agents that make judgments in law/tax/audit where wrong answers are unacceptable.
TR has 4,500 domain experts (largest employer of lawyers in the world), 1.5TB+ proprietary content, $200M/yr AI product investment, 200+ scientists/engineers.
Agency framed as four dials — autonomy, context, memory, coordination — tuned per use case rather than a binary agent/non-agent label.
Evaluation pain: same expert lawyers swing ±10% on identical questions a week later; trust requires near-deterministic behavior but agents are stochastic.
Legacy applications become reusable tools agents can invoke — decomposing 100-year-old software into agent-callable components is a moat.

agentshigh-stakesevals

Original description

This keynote will explore what it takes to move from basic generative assistants to fully agentic AI—systems that don’t just suggest but plan, act, and adapt—all within the structured, high-trust environments where professionals actually work.

About Joel Hron
Joel Hron is a passionate innovator driving the future of product technology and AI at Thomson Reuters. As Chief Technology Officer, he leads Product Engineering and AI Research & Development, pushing the boundaries of what’s possible in Legal, Tax, Audit, Trade, Compliance, and Risk solutions.

Joel joined Thomson Reuters in 2022 through the acquisition of ThoughtTrace, where he served as CTO. Previously, he led AI and TR Labs, launching seven groundbreaking GenAI products in just 18 months, transforming legal research, tax analysis, and contract drafting.

His approach is centered on rethinking processes through technology, building teams rooted in trust, transparency, and customer-centric innovation. Joel envisions AI not as a replacement for human expertise, but as a force that enhances professional decision-making, making expert information more accessible and impactful.

A New Orleans native, Joel’s global career spans work in London and Africa, and he now calls Zug, Switzerland home. He holds a Master’s in Mechanical Engineering from the University of Texas at Austin and a Bachelor’s in Engineering from Texas Christian University.

Recorded at the AI Engineer World's Fair in San Francisco. Stay up to date on our upcoming events and content by joining our newsletter here: https://www.ai.engineer/newsletter