| The same capabilities that could help develop vaccines, treatments and pandemic defenses could also be used to develop biological threats. Researchers and policy experts aren’t saying we’re at that point yet, because there are still significant barriers between an AI model and someone successfully creating a dangerous pathogen. But there is a growing concern about the trajectory and whether Washington is building safeguards fast enough. Let’s start with the technology. While there are overall concerns about whether AI could help someone create a biological weapon at some point in the future, experts are more troubled by what could happen if the technology is able to do more of the scientific work itself. That’s the focus of a recent report from WP Intelligence Lead Health Care Analyst Rebecca Adams. For example, a scientist from the Massachusetts Institute of Technology testing the bounds of an AI model found that it had combined information from different areas of biology into what appeared to be a road map for a new type of bioweapon. Although the company in question — which the scientist would not identify — and others in the industry have put filters in place that aim to decrease the likelihood that a user could get an AI model to produce dangerous information, there’s an unease that those protections may not go far enough. And, what’s more, some of the leading Chinese AI models don’t have the same safeguards. In another experiment, different researchers used AI to design 16 new viruses capable of attacking bacteria. While just a small share of the AI-generated sequences worked in the lab — and the viruses could not infect humans, only bacteria — the experiment offered a glimpse of where the technology could go. “It’s arguably unwise to release increasingly capable models to the public without guardrails that block off most of molecular and cellular biology by default. They will be able to reason about the existence of weapons that humans had not considered,” said Kevin Esvelt, the MIT associate professor who had worked to ferret out security risks in one tech company’s large language model (LLM). Why it matters: The advancements are coming as the Trump administration’s biosecurity oversight capabilities are weakened, and there’s still a lack of protections in place as the federal government has largely — until more recently — leaned into taking a hands-off approach. There are also some modest provisions in the defense authorization bill to strengthen AI-driven biosecurity, but experts say that users may be able to circumvent them. Read the full report: “Washington’s next test: The rise of AI-assisted bioterror.” EXECUTIVE HEALTH BRIEF Now, let’s pivot to Washington’s policy-development problem. The challenge of crafting guardrails around this rapidly evolving technology is unfolding as the federal government has lost some of the expertise and infrastructure that was supposed to help it think through biological threats. And the fear that AI could help unleash deadly pathogens has spurred the Trump administration to strengthen the federal government’s biosecurity policies, The Post’s Ian Duncan recently reported. It comes after Trump dismantled much of the White House infrastructure that handled the issue during the previous administration. At the end of the Biden administration, the White House had as many as 30 people working on biological security, coordinating across agencies and advising senior officials. However, after President Donald Trump returned to office, that team rapidly dwindled amid broader staffing cuts and reorganizations. At times, former officials told The Post, there was nobody at the White House working solely on biosecurity. In place of the more comprehensive team, the White House has leaned on a smaller, revolving set of people who either work part-time or have limited experience in the field, Ian reports. The former officials spoke on the condition of anonymity to avoid professional retaliation. “They have kneecapped themselves in their ability to create any policy that would have staying power,” one of the former officials, who recently worked directly on outbreak-response policy, told Ian. “That’s the risk, they’re not playing the long game.” Compounding the coordination vacuum, there are other offices or teams working on health security, biological threats, or pandemic preparedness throughout the government that have also been shuttered or reshuffled as part of a larger reorganization. While there is work throughout the government — including the Pentagon — to develop new defenses against future pandemics and deter biowarfare, analysts and former federal experts told The Post that it’s all unfolding while the federal workforce is reeling from broader brain drain and overall cuts to federal public health and science funding. Now, the Trump administration says it is working to replace Biden-era policies governing AI and biological risks, which the White House has criticized as inadequate. Trump revoked an executive order from 2023 calling for stricter biological safeguards, and directed officials to replace a separate policy designed to require companies to screen buyers seeking genetic components that could be used to create biological agents. That replacement policy was supposed to be produced by last August. It still hasn’t arrived. There are signs the administration is moving again. The White House has held briefings with OpenAI and with other government agencies on biological risks, and a new initiative involving the White House, the Department of Health and Human Services and others is reportedly taking shape. (The Washington Post has a content partnership with OpenAI.) “The Administration’s commitment to addressing AI and biological safety risks is clear and ongoing,” a White House official told Ian in an email. “The White House has been developing policy for over a year, including in our AI Action Plan last summer.” Read the full story: “Trump aims to rebuild defenses against bioweapons as AI fears rise.” “2 measles-related deaths reported in Pennsylvania, the state’s first in 35 years,” The Post’s Lena H. Sun and Rachel Roubein report. “More US employers to drop weight-loss drugs in 2027, as healthcare costs increase,” Amina Niasse writes at Reuters. “FTC’s Deals on Drug Price Reform Have Critics Questioning Scope,” report Bloomberg Law’s Lauren Clason and Katie Arcieri. This newsletter is published by WP Intelligence, The Washington Post’s subscription service for professionals that provides business, policy and thought leaders with actionable insights. WP Intelligence operates independently from The Washington Post newsroom. Learn more about WP Intelligence. | 
