The U.S. Food and Drug Administration has taken nearly five years to decide some applications for lower-risk tobacco products, and artificial intelligence tools could cut that wait by handling routine tasks, according to a report published by the R Street Institute in 2026. The analysis argues that while speed can't trump scientific rigor, tested AI systems could handle repetitive work like checking whether files are complete, finding specific evidence in thousands of pages, and flagging conflicting data—freeing human reviewers to focus on whether a product meets public-health standards. The report says faster decisions matter because combustible cigarettes cause more than 480,000 deaths each year in the United States, and adults who switch completely to lower-risk alternatives may reduce their health risks.
The FDA's Center for Tobacco Products took nearly five years to decide applications for ZYN nicotine pouches, nearly four years to authorize the Vuse Alto e-cigarette, and more than three and a half years to decide the original IQOS heated-tobacco modified-risk applications, according to the report. In May 2026, the center reported it had cut its application backlog by about 70 percent during 2025 and cleared its acceptance-review queue entirely. When a 2020 deadline required manufacturers of every e-cigarette already on the market to file a premarket application or pull the product, the center received 2,151 applications covering more than 6.6 million e-cigarette products. By October 2022, about 0.8 percent of those products remained pending, most had been rejected at an early stage or denied, and only 23 products had received marketing orders.
The report finds that three problems slow tobacco-product review: the sheer volume of applications, the complexity of the science they contain, and gaps in process-related guidance for submissions. According to the authors, some applications reportedly exceeded 125,000 and 150,000 pages, making it difficult for reviewers to find and connect facts scattered throughout. An independent review called the center's process cumbersome, unpredictable, and lacking transparency, the report notes, and found that guidance often came late and that the FDA's information systems weren't built to support a single, integrated application-review process. The report states that AI should never decide whether to approve or deny an application but should handle bounded tasks that produce outputs a scientist can check.
The report explains that rule-based checking tools can catch application defects like missing required fields and mismatched product names, while AI-based search and extraction tools could help reviewers find and organize specific facts with high accuracy when the task is narrow and the source documents follow familiar patterns. The FDA completed an AI-assisted scientific-review pilot in 2025 and launched Elsa, an internal generative AI assistant, across FDA centers, which can summarize documents, compare labels, and analyze reports of possible health problems. One FDA official said AI tools cut the time needed for some tasks from three days to minutes, though the agency hasn't published enough detail to judge the pilot's error rate or its effect on overall review time. The report cautions that AI can invent citations, omit important evidence, repeat past inconsistencies, or encourage "automation bias"—the tendency to trust a computer answer too readily—and that any of these errors could expose the agency to legal challenge if reflected in a decision.
The report recommends that the CTP create a formal AI-assisted review program, require key application data be submitted in standard electronic formats, and test narrow uses like completeness checks and citation verification before expanding to more demanding tasks. The authors argue that humans must continue to make every high-stakes decision, including judgments about study credibility and whether a product meets public-health standards, and that applicants should be told when AI materially affects a review. The report concludes that well-governed AI can reduce clerical work and organize evidence without replacing scientific and legal judgment, helping the center reach sound decisions sooner so that products meeting the law's public-health standard can reach adults who smoke with less avoidable delay.