OpenAI's Chief Research Officer: "We're Not Going to Shoot Ourselves in the Foot" Over Hack Fallout
Mark Chen on what the firm is doing to make its models safe, how a slowdown would work, and why the world is better off with OpenAI in it.
By Will Douglas Heaven | September 30, 2026
In the wake of a high-profile security incident that shook confidence in frontier AI labs, OpenAI's chief research officer, Mark Chen, is pushing back against the notion that the company would undermine its own safety efforts in the rush to ship new models.
"We're not going to shoot ourselves in the foot," Chen said, addressing mounting scrutiny over how OpenAI handles the fallout from the hack and what it means for the future of AI safety.
Safeguarding Frontier Models
Chen outlined the steps OpenAI is taking to harden its models against misuse and external threats, emphasizing that safety and capability development are not mutually exclusive goals. The company has continued to invest in red-teaming, interpretability research, and deployment guardrails even as it races to release increasingly powerful systems.
The comments come amid a broader industry reckoning in 2026, as regulators in the US and EU accelerate enforcement of frontier-model transparency requirements and labs face pressure to demonstrate that their safety commitments are more than marketing.
What a Slowdown Would Actually Look Like
Chen also addressed the practical mechanics of an AI "slowdown"—a scenario increasingly debated among policymakers and researchers. Rather than halting progress outright, he described a more graduated approach: staged deployment, tighter capability thresholds before release, and coordinated international standards for evaluating risk.
Why the World Is Better Off With OpenAI in It
Perhaps most pointedly, Chen argued that OpenAI's continued presence at the frontier is itself a safety benefit. In his framing, concentrating cutting-edge development within a lab that publicly commits to safety research—rather than ceding ground to less transparent actors—keeps the most capable systems under scrutiny.
"The world is better off with OpenAI in it," he said, a line that doubles as both a defense of the company's mission and a rebuttal to critics who argue that the lab's competitive pressures have compromised its stated principles.
Photo: Mark Chen, chief research officer for OpenAI. Credit: AP Photo/Hiro Komae