THE SIGNAL
Goldman Sachs spent six months working with Anthropic engineers to build autonomous agents for two specific functions: trade reconciliation and client onboarding. This week the story got bigger. Goldman, Anthropic, Blackstone, and Hellman & Friedman announced a joint enterprise AI services firm designed to bring Claude into core operations across all four companies. Goldman named this direction in its Q4 2025 earnings call as "One Goldman Sachs 3.0," a firmwide operating model rebuilt from the ground up around AI.
WHY IT MATTERS TO YOU
The workflows Goldman automated first are the ones early analysts spend their first year doing: trade-reconciliation checks, document review, onboarding paperwork, and rule-based compliance flagging.
Goldman still posted 2026 summer analyst and compliance analyst positions this month. Entry-level doors are open. What is changing is the job description before you arrive. The analysts who move fastest will be the ones who can supervise the agent, catch its errors, and explain its outputs to someone senior.
THE SKILL OR TOOL
In February, I completed Goldman's Operations simulation on Forage, which runs you through trade settlement discrepancies and HNW asset transfers, the exact workflows now being rebuilt around Anthropic agents. That simulation is still the most direct window into what Goldman expects from entry-level operations and compliance candidates. Pairing it with Anthropic's AI Fluency certification gives you the domain context and the tool fluency in one credential stack, and both are free. Goldman Operations is at theforage.com, search Goldman Sachs. AI Fluency is at anthropic.com/academy.
THE QUESTION
If the agent handles reconciliation and onboarding, what is the one skill you are building right now that cannot be replaced?
— Marco Meneses Finance & Business Analytics @ Wilkes University | Greenwood Project FinTech Scholar theanalystedgehq.com
Reading this on LinkedIn? Get it in your inbox every Thursday at theanalystedgehq.com