Thomson Reuters and Snowflake define fiduciary grade AI for legal and tax professionals

Thomson Reuters builds fiduciary grade AI on governed authoritative data

Thomson Reuters is demonstrating how fiduciary grade AI operates in practice by building its enterprise AI and data platform on Snowflake to serve legal professionals, tax professionals, and accountants who carry personal liability for their decisions 1

1

. The company's data estate spans more than 37,500 governed tables and 350 databases, providing the foundation for AI tools that professionals can stake their reputations on in high-stakes environments 2

2

. According to Bala Kasiviswanathan, vice president of developer and AI experiences at Snowflake, "All these tools for AI are not real until they are on a governed data platform" 1

1

.

Four pillars define trusted and governed enterprise AI standards

For Thomson Reuters, fiduciary grade AI represents an operational standard built on four components: authoritative content, investment in data security and safety, expert validation through subject matter experts who vet outputs, and transparent results 1

1

. Laura Safdie, head of legal innovation at Thomson Reuters, explained the stakes clearly: "The difference between using AI that's unconnected to legal data, unconnected to legal expertise can mean getting it wrong. And so we work with the professionals who need to get it right every time" 1

1

. This standard ensures that professionals facing personal liability can trust AI outputs in their critical decision-making processes.

CoCounsel evolves into agentic legal AI assistant on authoritative foundation

CoCounsel, Thomson Reuters' legal AI assistant co-created by Safdie before the company acquired Casetext in 2023 for $650 million, exemplifies how fiduciary grade AI delivers trusted professional insights 1

1

. The assistant has evolved from an early retrieval-augmented generation approach into a fully agentic assistant natively built on Thomson Reuters' authoritative legal content, including Westlaw and Practical Law 1

1

. Safdie noted that rebuilding around an agentic harness unlocks complex work products while keeping lawyers in control of every step, with Thomson Reuters' responsible AI team vetting every capability for hallucination and bias before deployment 1

1

.

Data governance accelerates enterprise AI rather than constraining innovation

Thomson Reuters selected Snowflake in 2021 for its ability to bridge enterprise-grade data governance with scalable infrastructure 2

2

. The company has created a single, secure source of truth that powers its internal My Data Space platform, where more than 1,500 internal users including data engineers, analysts, and business leaders access regulated data daily 2

2

. By building semantic capability on Snowflake that unifies more than 23 previously fragmented customer master data sources, Thomson Reuters has given finance teams a single, trusted definition of core business terms 1

1

. Kasiviswanathan emphasized that "self-serve is the future," noting that a secure governed platform removes bottlenecks and accelerates enterprise AI deployment 1

1

.

Snowflake Cortex AI and CoCo drive measurable performance gains

Leveraging Snowflake Cortex AI, Thomson Reuters is moving beyond data management to deliver AI at scale across the organization 2

2

. Caitlin Halferty, head of data and analytics at Thomson Reuters, stated: "With Snowflake Cortex, we're accelerating how we build and scale AI across Thomson Reuters. The real value is not just speed. It is the ability to innovate in a governed environment where our teams can turn complex regulatory data into actionable insights" 2

2

. As the company consolidates data pipelines supporting flagship products like CoCounsel and Westlaw, key workloads are running up to 3.4 times faster, enabling teams to move from static reporting to near real-time insights . Complex analysis that previously took weeks now takes seconds, with manual data preparation eliminated across key workflows .