OpenAI’s head of data centers has left after a short tenure, adding a closely watched infrastructure position to the company’s growing list of executive departures. Chris Malone joined in March 2025 and departed last week, according to a Wall Street Journal report cited by TechCrunch.
Malone previously spent almost five years at Meta and more than a decade at Google. His exit is notable because OpenAI is pursuing a vast physical expansion at the same time that leadership responsibilities inside its infrastructure organization are being rearranged.
A critical role changes during a buildout
Malone arrived soon after the launch of the Stargate Project, a US data-center initiative associated with OpenAI, Oracle, Nvidia, SoftBank and Microsoft. The program makes execution across land, power, construction, computing systems and delivery schedules central to OpenAI’s ability to expand its services.
The reason for Malone’s departure is unclear. OpenAI told TechCrunch that it had recently reorganized the infrastructure organization to support the scale and pace of its work. The company said it retains a deeply experienced data-center team with clear leadership and technical expertise. That statement describes OpenAI’s confidence in the group; it does not independently establish how the transition affects individual projects.
Before leaving, Malone stopped reporting directly to OpenAI president Greg Brockman and began reporting to vice president Sachin Katti, who assumed leadership of the wider group, according to the Journal. A reporting-line change can be a routine response to scale, but its proximity to a senior departure makes the reorganization part of the story.
Responsibility is distributed across several leaders
OpenAI still has multiple executives overseeing different parts of its infrastructure effort. Uday Ruddarraju leads the data-center team, Brent Mayo leads the build and delivery program, and Spas Lazarov leads data-center engineering. Lazarov brings experience from the data-center and energy sectors.
That division suggests the company is separating strategy, construction delivery and engineering rather than placing every responsibility under one executive. Distributed ownership can add specialist depth, although it also requires precise coordination when schedules, power needs and technical design depend on one another.
The source does not identify missed targets, project cancellations or operational failures connected to Malone. It also does not say whether a direct replacement will be appointed. Any conclusion that the departure has already disrupted OpenAI’s expansion would therefore go beyond the available evidence.
Executive churn adds pressure before a listing
Malone joins more than a dozen OpenAI executives reported to have left during 2026. Recent changes include the replacement of chief revenue officer Denise Dresser after roughly eight months, the departure of long-serving chief operating officer Brad Lightcap and Fidji Simo’s move from product and business chief to an advisory role while addressing a chronic illness.
The company has also lost ethics and safety personnel, disbanded its preparedness team, and seen leaders leave after projects were closed. Greg Brockman has argued that OpenAI’s visibility causes every departure to receive unusual scrutiny. That may explain some attention, but the count and seniority make the pattern relevant to governance and execution.
TechCrunch reports that an OpenAI public listing once expected in 2026 has been pushed to 2027. Questions about valuation, profitability and large capital requirements are already part of that preparation. A leadership change in data centers intersects directly with those concerns because infrastructure converts financial commitments into operating capacity.
The immediate facts remain limited: Malone has left, reporting lines were reorganized and several experienced leaders remain. Investors and partners will need evidence from construction milestones and service capacity to judge the effect. Until then, the departure is a governance signal rather than proof that OpenAI’s infrastructure plan has succeeded or failed.
