Ben Buchanan

Ben Buchanan received a Ph.D. in war studies from King's College London, where he was a Marshall Scholar, and master's and undergraduate degrees from Georgetown University. At Georgetown he was director of the CyberAI Project and a senior faculty fellow at the Center for Security and Emerging Technology (CSET), and an assistant teaching professor at the School of Foreign Service, researching "the intersection of cybersecurity and statecraft." He wrote The Cybersecurity Dilemma (Oxford University Press, 2017) and The Hacker and the State: Cyber Attacks and the New Normal of Geopolitics (Harvard University Press, 2020), contributed regularly to War on the Rocks and Lawfare, and published op-eds in The Washington Post.¹ As a CSET senior faculty fellow he testified in October 2019 before the House Homeland Security Subcommittee on Cybersecurity, Infrastructure Protection, and Innovation, at its hearing "Preparing for the Future: An Assessment of Emerging Cyber Threats." His CSET work of 2020 and 2021 included "The AI Triad and What It Means for National Security Strategy" (August 2020), "Automating Cyber Attacks" (November 2020) and "Truth, Lies, and Automation" (May 2021), a study of whether natural language models such as OpenAI's GPT-3 could be used to generate automated disinformation at scale.¹

White House

According to documents the Washington Examiner obtained under the Freedom of Information Act, CSET's founding director, Jason Matheny, while serving in senior White House roles between March 2021 and June 2022, hired Buchanan, "then a top staffer at CSET," under the Intergovernmental Personnel Act and tasked him with helping the administration craft its AI policies. Georgetown continued to pay Buchanan's salary and benefits as director of the CyberAI Project, and the federal government reimbursed them at the end of each month. The paper wrote that CSET "was funded almost entirely by Open Philanthropy, receiving $97 million from the charitable foundation between 2019 and 2021," that Open Philanthropy was funded primarily by Dustin Moskovitz and Cari Tuna, and that Moskovitz "was a major investor in Anthropic's initial $124 million fundraising round, which occurred in early 2021," around the time Matheny and Buchanan entered the administration.²

Buchanan worked in the White House from 2021 to 2025, on the National Security Council and as the first White House Special Advisor for Artificial Intelligence, "responsible for coordinating the Biden Administration actions on AI."³⁴ He worked on National Security Memorandum 25, which in his words "aimed to catalyze the use of AI for national security purposes and to prevent strategic surprise to the United States."⁴ The Washington Examiner, citing Semafor, described him as "one of the key architects" of the Biden administration's restrictions on exports of advanced chips to China. Shortly after Donald Trump took office in 2025, Anthropic wrote to his administration urging it to keep the export controls in place; the Trump administration rescinded the rule.²

Asked by the Washington Examiner about the arrangement, Buchanan said: "For my entire career, I have worked to ensure that the United States," not China, "leads the world in AI and that the technology is safe, secure, and trustworthy. It was an honor to take a leave from my professorship at Georgetown University to advance this mission in the White House. I scrupulously complied with all legal and ethics rules and answered only to President Biden and the senior White House staff. Congress passed the Intergovernmental Personnel Act to bring technical academics into government to work on pressing issues." He added: "Both President Biden and President Trump have used this authority to hire AI experts, and I was proud to be one of them."²

Johns Hopkins and Anthropic

Buchanan became the Dmitri Alperovitch Assistant Professor at the Johns Hopkins University School of Advanced International Studies in Washington.³ On April 7, 2025, at the university's Emerging Technologies symposium, he defended the AI diffusion rule, issued in the last days of the Biden administration, which sorted countries into tiers for access to advanced chips and placed the United Arab Emirates, Saudi Arabia and India in the second tier. Microsoft and Nvidia were lobbying the Trump administration to reverse it. "If computing power is so fundamental to AI, and the US and our allies are at such an extraordinary advantage in the design, production and use of that computing power, then we should be careful where we send this," he said. Of the companies' objection, he said: "The argument was that we were limiting who these US companies can sell these chips to and therefore limiting the revenue of these companies. Nvidia's stock, prior to the tariffs at least, did very well and they've done just fine, because there's extraordinary demand for AI chips."⁵

On July 21, 2026, testifying before the House Permanent Select Committee on Intelligence at the hearing "25 Years After 9/11: Confronting the Next Generation of Threats," Buchanan described himself as "a professor at Johns Hopkins University, the author of four books on cybersecurity and AI, and an outside advisor to AI and cybersecurity companies, such as Anthropic." He told the committee that the United States "must continue to take actions, such as denying advanced chips and chipmaking equipment to China, that ensure this lead endures." As evidence of AI's role in cyber operations he cited Anthropic's report of November 2025 that "a Chinese state-sponsored group had used its Claude model to run a largely automated cyber-espionage campaign against roughly thirty companies and government agencies across several countries, with the AI carrying out an estimated 80-90 percent of the tactical operation on its own," a Google report that Russian military intelligence had deployed malware in Ukraine that "uses AI mid-operation to write its own attack commands," and Anthropic's August 2025 exposure of a criminal extortion operation. He cited Anthropic's statement that "the vast majority of its production code is written by AI but overseen by humans," and told the committee that AI "will also raise new and grave concerns in areas as diverse as cybersecurity, biology, and accidental loss of control."⁴