FDA Launches Its AI Tool 'Elsa' Ahead of Schedule to Enhance Agency Operations
The FDA says its new large language model–powered tool is already accelerating reviews, aiding inspections, and supporting database development as part of FDA’s broader AI modernization strategy.
With the flurry of FDA updates, we’ll be sharing brief, semi-regular news roundups — quick, digestible explainers to keep you on top of the latest developments. Make sure you’re on our paid tier to get each issue in full.
The FDA has officially launched Elsa, a new agency-wide AI tool designed to modernize operations and improve efficiency across its regulatory functions. The launch, announced on June 2 by FDA Commissioner Marty Makary, came nearly a month ahead of the agency’s original June 30 rollout target, following a successful pilot program with scientific reviewers.
“Elsa is already transforming how we work,” said Makary. “Tasks that previously took days can now be completed in minutes. One scientific reviewer told me what once required two to three days now takes six minutes.”
The agency states that it has started utilizing the tool to expedite clinical protocol reviews, streamline safety profile assessments, facilitate label comparisons, and support database development for nonclinical applications. Additionally, Elsa is aiding inspectors in identifying high-priority inspection targets.
What we know about Elsa
Elsa is a large language model (LLM)–powered AI tool built within the FDA’s secure GovCloud environment. It is designed to assist employees with reading, writing, summarizing, and more specialized functions such as summarizing adverse events, performing expedited label comparisons, generating code for nonclinical applications, and helping identify high-priority inspection targets. The tool’s models do not train on data submitted by the regulated industry, and all processed information remains securely housed within the agency.
“AI is no longer a distant promise but a dynamic force enhancing and optimizing the performance and potential of every employee,” said FDA Chief AI Officer Jeremy Walsh. He described Elsa’s launch as the first step in a broader AI journey that will see the technology integrated across additional FDA processes, including data processing and generative-AI functions, to further support the agency’s mission.
A milestone launch amid budget and staffing headwinds
Elsa’s debut comes at a challenging time for the agency. The Trump administration’s proposed FY2026 budget would cut FDA funding by approximately 4%, reducing its budget to $6.8 billion and eliminating an estimated 1,940 staff positions. In April, the FDA laid off 3,500 employees—including scientists and inspection staff—though some of those cuts were later reversed. Staff reductions and missed PDUFA deadlines this year have already strained agency resources.
Against this backdrop, Elsa is being framed as a tool that could help offset lost capacity and improve the speed of regulatory reviews. In a May press release, Makary said he was “blown away” by the tool’s potential to accelerate new therapy reviews. “We need to value our scientists’ time and reduce the amount of non-productive busywork that has historically consumed much of the review process,” he added.
Independent experts agree that if implemented safely, AI could deliver meaningful efficiency gains. “The rollout of Elsa is a strong step forward by the agency in adopting AI to improve the speed and cost-effectiveness of drug development,” said Jon Walsh, Ph.D., Chief Scientific Officer of clinical trial AI firm Unlearn. “Saving time in the drug review process while maintaining high standards for scientific rigor is a big win for patients.”
Caution and concerns around “rushed” rollout and AI “hallucinations”
Despite enthusiasm around Elsa, concerns about the tool’s readiness and accuracy have emerged in the days following its launch.
According to Ars Technica reporting, FDA staff have described the rollout as “rushed” and raised concerns that Elsa’s capabilities have been overstated by agency leadership. Ars reports that during internal testing, FDA staff encountered inaccurate and incomplete answers when querying Elsa about FDA-approved products and other public information—suggesting that the tool may not yet be ready for use in scientific review contexts.
Sources speaking to STAT News and NBC News voiced similar concerns, warning that Elsa currently lacks sufficient guardrails and should be limited to administrative tasks for now, rather than regulatory or scientific decision-making. One staffer quoted by Ars said, “Makary and DOGE [Department of Government Efficiency] think AI can replace staff and cut review times, but it decidedly cannot.”
The reporting also notes that Elsa is based on Anthropic’s Claude LLM and was developed under a contract with Deloitte, which is now scaling the tool across the agency. Previously, individual FDA centers were developing their own AI pilots—such as CDER-GPT—before leadership chose to unify the effort under the Elsa brand.
The risk of AI “hallucinations”—false or misleading outputs—remains an industry-wide challenge. Federal IT experts interviewed by Gizmodo note that such hallucinations can stem from biases in training data or gaps in model safeguards. “Human oversight is essential to mitigate the risks and ensure the reliability of AI-integrated federal data streams,” one report cautioned.
Looking ahead: Elsa’s role in the FDA’s AI roadmap
Time will tell how Elsa performs in practice and how much it ultimately impacts review timelines, especially as the FDA works to recover from recent staffing and funding challenges. Since January, the agency has already missed several key PDUFA dates, and it remains to be seen whether AI can meaningfully improve throughput without compromising review quality.
For now, the FDA is moving forward cautiously, offering staff training on Elsa and planning an incremental expansion of its capabilities. “As we learn how employees are using the tool, our development team will be able to add capabilities and grow with the needs of employees and the agency,” said Walsh.
What should RA/QA teams take away from Elsa's launch?
While Elsa is an internal FDA tool with no direct industry-facing components (yet), it signals an important shift in how the agency will approach regulatory review and inspection work moving forward.
A few things to start thinking about here:
Expect more potential variability in review timelines. The FDA is positioning Elsa to accelerate certain parts of the review process—particularly clinical protocol reviews and adverse event summarization. However, early reports suggest variability in how different centers, divisions, and reviewers are using the tool. Teams may be wise to be prepared for some inconsistency in review pace and depth over the next 6–12 months as FDA adoption normalizes. Where the FDA begins to leverage Elsa, we may expect to see more rapid turnaround on data-heavy submissions (e.g., adverse event reports, labeling comparisons).
Anticipate AI-assisted inspections. Elsa is being used to help inspectors identify “high-priority inspection targets.” This suggests that risk-based inspection targeting will increasingly rely on AI-derived signals. RA/QA teams should review their inspection readiness assumptions now. The FDA may now pick up on patterns or signals that weren’t part of prior risk models.
Prepare for a potential shift in reviewer expectations. As reviewers grow accustomed to Elsa summarizing and synthesizing data, expectations around clarity, structure, and “AI-friendly” formatting of submissions may evolve. Well-organized, machine-readable documents could give sponsors an edge, reducing the risk of misinterpretation or errors during AI-assisted review. Consider collaborating with regulatory writers and data teams (many of whom we deploy on a contract basis) to ensure that submission quality and consistency remain high.
Scroll through all our recent FDA news updates here.
Who is The FDA Group?
The FDA Group helps life science organizations rapidly access the industry's best consultants, contractors, and candidates. Our resources assist in every stage of the product lifecycle, from clinical development to commercialization, with a focus on Quality Assurance, Regulatory Affairs, and Clinical Operations.
With thousands of resources worldwide, hundreds of whom are former FDA, we meet your precise resourcing needs through a fast, convenient talent selection process supported by a Total Quality Guarantee. Learn more and schedule a call with us to see if we’re a fit to help you access specialized professionals and execute your projects on time and on budget.