Goldman Sachs maps 10 AI use cases as top RIAs swell to $1.6T

At the Goldman Sachs Annual RIA Professional Investor Forum in New York on Tuesday, executives presented an AI plan focused on 10 use cases and released data showing the top 100 registered investment advisor firms now manage more than $1.6 trillion in client assets, double the amount from two years ago. Padi Raphael, global co-head of third party wealth, reported audience size for those firms grew about 15% over the same period.

John Waldron, Goldman Sachs president and chief operating officer, outlined a ‘one Goldman Sachs’ architecture that concentrates work on a limited set of initiatives. The bank is focusing resources on lending processes, client onboarding, data infrastructure, trade ledgers, liquidity platforms and software development and holds weekly reviews to track returns.

Goldman reported early results from AI coding tools that show software developers achieving 20-50% productivity gains. The firm is not reducing headcount as it adopts AI; instead it is keeping staff levels while completing more projects with the same number of developers and lowering its zero-based budget line.

The bank measures what it calls gross benefit across three categories: productivity gains from existing staff, hard cost savings from process automation, and foregone investment when planned hiring becomes unnecessary. Waldron advised against running too many AI projects at once and recommended scaling front, middle and back office functions together. He suggested dividing large organizations into smaller community units to preserve culture rather than operating as a single large unit.

Raphael said the concentration of assets in fewer, larger RIA firms creates operational challenges tied to scale, technology deployment and talent management.

Waldron described AI as currently inflationary, noting compute infrastructure costs are rising between 10% and 30% and that data center and specialized chip costs are increasing. He said capital expenditures related to AI are likely to climb over the next 12 to 24 months before efficiency gains materialize.

On markets and dealmaking, Waldron said the current M&A cycle is being led by corporate buyers seeking scale to fund AI investments and capture margin advantages, and that larger firms are earning premium valuations. He reported continued U.S. equity dominance in many investor portfolios and characterized the economy as K-shaped, with lower-income consumers under pressure while broader indicators remain firm.

Goldman described its hiring process as multi-stage, testing skills in early rounds and focusing on cultural fit in later rounds to protect organizational culture during rapid scaling. On capital markets, Waldron said private credit has a place in client portfolios but called for better disclosure and industry discussion about suitability and liquidity. He added that large liquidity events expected from venture-backed companies in 2026 and 2027 could return capital to private markets.

Waldron urged privately held RIAs to stay private longer to avoid short-term public pressures, stating, ‘This quarter-to-quarter thing is challenging when you’re trying to build enduring value.’ Forum presenters emphasized a focused, disciplined approach to AI deployment and the need to measure returns on a limited set of projects.