OpenAI security chief Matt Knight steps down after five years; Sam Altman thanks him publicly

OpenAI

• Matt Knight, OpenAI's Chief Information Security Officer, resigns after 5 years
• Knight led security initiatives including the launch of Aardvark at OpenAI
• OpenAI CEO Sam Altman praised Knight's impact and leadership in security

Did our AI summary help?

Matt Knight, Vice President and Chief Information Security Officer at OpenAI, has announced his resignation after more than five years at the company. Knight shared the news in a detailed post on X, where he published a note he said had already been circulated internally at OpenAI. The departure marks the exit of one of the company’s earliest senior leaders in security, at a time when OpenAI’s platforms are used by hundreds of millions of people worldwide.

In his note, Knight reflected on joining OpenAI in 2020 as its first dedicated security hire. At the time, the organisation was still a relatively small research lab, best known for launching GPT-3 and experimenting with an early API. Since then, OpenAI has grown into one of the most influential players in artificial intelligence, with products that sit at the centre of global debates around safety, trust, and responsible deployment.

Knight said he was proud of how OpenAI’s security programme evolved during his tenure. He described helping build systems that now protect a global platform trusted by hundreds of millions of users and safeguard some of the most important AI systems currently in operation. According to Knight, a central focus of his work was proving that security could keep pace with fast-moving research, rather than slowing it down.

He also highlighted a more recent phase of his role, during which he led security product development and research. That work culminated in the launch of Aardvark, a security initiative that Knight described as a reflection of his core beliefs about how artificial intelligence can meaningfully improve software defence. He said bringing Aardvark into the world was among the most fulfilling chapters of his career, both technically and personally.

Knight explained that with Aardvark now launched and the broader security programme in a strong position, the timing felt right for him to move on. He said he was leaving confident that both the security organisation and the product teams were in capable hands, and that the work he helped start would continue to grow and have impact.

Much of Knight’s message was devoted to thanking colleagues. He described security as a team effort and expressed gratitude to those who believed in OpenAI’s security mission when it was still largely theoretical. He paid particular tribute to members of the security organisation, past and present, saying their rigour, creativity, and care made the team’s achievements possible.

OpenAI CEO Sam Altman responded publicly to Knight’s post with a message of appreciation. Altman thanked Knight for his contributions, from being OpenAI’s first security hire to building what he described as a world-class security team and launching Aardvark. He said Knight’s impact on the company had been enormous and wished him well in whatever he chooses to do next.

Knight did not disclose his next move, saying only that he is looking forward to exploring new ideas and opportunities. He invited people to stay in touch and signed off his note on an optimistic tone, encouraging others to stay curious.

Knight’s departure comes as OpenAI continues to face increasing scrutiny over security, trust, and governance as its tools become more deeply embedded in consumer, enterprise, and government use cases. His exit closes an important chapter in the company’s evolution from a small research lab into a global AI platform, and places renewed attention on how OpenAI will maintain and evolve its security posture in the years ahead.