Thomas Lind, the head of policy at the Office of the National Cyber Director and a senior adviser to National Cyber Director Sean Cairncross, plans to leave government soon, the latest in a string of exits from the White House's technology policy ranks just days after President Donald Trump signed a sweeping AI executive order, Politico reported.
Two people familiar with the decision told the outlet Lind intends to depart to spend more time with family. He did not respond to requests for comment.
Lind's planned exit lands at a particularly awkward moment. The executive order Trump signed last Tuesday lays out actions that ONCD and other federal agencies must take in the coming weeks to fortify networks against hacking risks posed by frontier AI models, including Anthropic's Claude Mythos and OpenAI's GPT-5.5. Among the most consequential provisions: a voluntary safety-vetting system under which developers like OpenAI and Anthropic would submit new models for federal review 30 days before release. ONCD has 60 days to help stand that system up.
The office now has to do it with a shrinking bench.
Lind is not the first senior figure to head for the door. Alexandra Seymour, his former deputy on policy, exited last week. Sriram Krishnan, the White House's senior policy adviser for AI, plans to leave later this month. Politico reported that others from the White House cyber policy shop are also looking to exit.
The wave of departures follows what the outlet described as low morale inside ONCD, citing an industry official and a person familiar with the office's internal dynamics. Some administration officials blamed early setbacks surrounding the executive order on Cairncross, saying he did not move quickly enough to address concerns and failed to hire enough technical talent.
That staffing gap is real and measurable. ONCD is authorized to employ up to 75 people. Cairncross has staffed it with only about three dozen, according to Politico's prior reporting. A White House official said the office is "actively" looking to hire qualified candidates capable of advancing Trump's agenda.
The administration has seen high-profile departures from other agencies in recent months, but the concentration of exits inside the cyber director's office raises a distinct concern: the people who helped write the new AI rules are leaving before those rules take effect.
Lind was not a figurehead. A third person familiar with the process told Politico that Lind was involved in crafting the executive action and participated in meetings with Google, Anthropic, and OpenAI as those companies provided feedback to ONCD. He was described as one of the most technically savvy senior policy officials at the office.
His background bears that out. Lind left the NSA last April to become the White House cyber office's senior director for intelligence, according to his LinkedIn profile. He became an adviser to Cairncross in September and assumed the senior-most policy position at ONCD in April.
That trajectory, from the intelligence community to a policy leadership seat, is exactly the kind of technical expertise the office was designed to attract. Congress created ONCD in 2021 to oversee budgets and cyber defense policies across a sprawling federal executive. The national cyber director is supposed to serve as the principal adviser to the president on deterring malicious cyber activity and shoring up U.S. cyber defenses.
Losing the person who held the top policy role, barely two months after he took it, is not a minor staffing hiccup. It is a gap at the center of an office that just received a major new mandate.
Cairncross was confirmed in August 2025. Before that, he was a lawyer and Republican National Committee executive, a background that did not include cybersecurity or digital policy. That lack of technical grounding has drawn criticism from inside the administration, with some officials saying he moved too slowly on the executive order and failed to build out ONCD's roster.
The executive order itself had a rocky path to the president's desk. Last-minute opposition from the AI industry led Trump to scuttle an earlier draft. When the final version was signed last Tuesday, Politico noted it arrived "to little fanfare." The directive marked a shift from what the outlet characterized as the administration's previously lighter-touch approach to AI regulation, a shift that kicked into high gear in early April after Anthropic announced it had built a new model adept at finding holes in digital infrastructure.
Personnel turnover in any administration is normal. Senior officials step down for all kinds of reasons, and Lind's stated motivation, family, is as common as they come. But the timing and clustering of these exits matter. When three senior tech policy hands leave within weeks of a major order, and the office responsible for implementing that order is running at less than half capacity, the question is not whether the departures are understandable. The question is whether the work gets done.
White House spokesperson Liz Huston pushed back on the idea that ONCD is struggling, praising Cairncross and the broader team.
"Sean Cairncross and his entire team at the Office of the National Cyber Director continue to do outstanding work to strengthen America's cyber and national security, protect critical infrastructure, and ensure the United States remains the global leader in AI innovation."
Huston added that the office is working "hand-in-hand with the private sector and other federal agencies" to implement the president's policy priorities.
That framing is fine as far as it goes. But "outstanding work" and a 60-day implementation clock do not square easily with a policy shop that keeps losing the people who built the policy. White House Chief of Staff Susie Wiles charged ONCD with leading the response to AI models capable of unearthing software flaws inside digital infrastructure. That is a serious mission. It requires serious, sustained staffing.
Meanwhile, the companies most affected by the new order, OpenAI and Anthropic, have so far released their most advanced models only to a select group of tech organizations and security researchers. The voluntary review framework ONCD must build will shape how and when those models reach a broader market. The stakes are not abstract.
Personnel churn inside the Trump administration is hardly unique to ONCD. Staffing shakeups have rippled through other agencies, and the nature of political appointments means turnover is baked in. But there is a difference between normal rotation and a pattern that leaves a critical office short-handed at the worst possible moment.
The open questions are straightforward. Who replaces Lind? Who replaces Seymour? Can ONCD recruit technical talent fast enough to meet its 60-day deadline? And will the voluntary safety-vetting system carry any weight if the office standing it up cannot hold onto its own people?
Cairncross has a narrow window. The executive order's implementation deadlines are measured in weeks, not months. Hiring in government is slow under the best circumstances. Doing it at an office running at roughly half strength, with morale reportedly low, is harder still.
The broader pattern inside Trump-aligned circles, aides cycling out of senior roles and into the private sector, is not inherently alarming. People leave government. But the AI executive order represents one of the most significant federal forays into technology oversight this administration has attempted. If the office tasked with executing it cannot keep the lights on, the order risks becoming a document that sounds good and does nothing.
Washington has no shortage of ambitious policy directives. What it lacks, more often than anyone in power wants to admit, is the follow-through to make them stick.