Goldman Sachs to remove explicit DEI language from board selection criteria

Goldman to strip DEI language from board selection criteria

Goldman Sachs is moving to remove explicit diversity, equity and inclusion (DEI) language from the criteria its board uses to evaluate director candidates, people familiar with the matter tell the Wall Street Journal. The change targets an appended reference to “other demographics” — a phrasing that refers to race, gender identity, ethnicity and sexual orientation — and would leave a more traditional diversity framework that highlights viewpoints, background and professional experience.

The reported revision follows a September request from the conservative National Legal and Policy Center and, according to sources, comes as part of a voluntary arrangement to head off a formal shareholder submission ahead of Goldman’s annual meeting. The bank already removes a diversity requirement in standards it applies to companies it takes public, and the latest move is framed internally as a governance recalibration rather than a broader retreat from inclusion initiatives.

The decision arrives amid a wider political push against DEI across U.S. government and corporate circles since the Trump administration’s executive orders to curtail certain DEI programs. Observers say the step could prompt scrutiny from investors, proxy advisers and civil-society groups that monitor environmental, social and governance (ESG) metrics, while supporters argue it refocuses nominations on skills and board independence. Goldman has not yet issued a public filing detailing the change and stakeholders are watching for formal policy language, potential revisions to nominating committee procedures, and any regulatory or shareholder responses.

Goldman analysts weigh AI’s impact on incumbent software

Separately, Goldman analysts are advising a selective approach to software amid AI-driven market shifts. Gabriela Borges says agentic AI functions more as an “intelligent layer” that depends on high-quality historical data held by systems of record, keeping companies such as SAP, Salesforce, Oracle and Workday strategically important. The firm is looking for proof that domain experience drives superior agentic outcomes and stable-to-improving fundamentals when recommending names in the software ecosystem.

Memory supply squeeze ripples into consumer and enterprise markets

Goldman researchers are also flagging a high-bandwidth memory bottleneck tied to surging AI data‑centre demand. Analysts including Katherine Murphy and Allen Chang revise global PC shipment forecasts lower for 2026–28, warn the crunch may persist about two years, and note shortages are beginning to appear in retail channels, with some consumer devices intermittently out of stock. The bank highlights that the constraints extend beyond chips to the wider data‑centre buildout supply chain.