The Past Is Prologue: Three Insights from 2025 That Will Impact 2026
A year ago, I launched FDA Matters (www.fdamatters.com) to help the FDA stakeholder community better understand the FDA. My analysis and commentaries have focused on the policy, organizational, and political challenges faced by the agency.
Here are some highlights from my 2025 columns that seem particularly pertinent as we begin a new year.
FDA’s Credibility Is Tied to Its Reputation for Science-Based Decisionmaking
“Science is a discipline. As such, it has rules that must be followed for policy and regulatory decisions to be meaningful, actionable, and effective. Data may be subject to interpretation and conclusions can vary, but the fundamentals of scientific thinking are always applicable.”
Thoughts to carry forward into 2026:
In urging HHS Secretary Kennedy to listen to experts at CDC (https://bit.ly/4nlFAAc), I articulated “six rules of scientific thinking" that are immutable:
· Anecdotes are not proof, and they cannot be generalized. At best, a collection of anecdotes serves as the basis for a scientific hypothesis.
· Correlation is not causation. Our brains are wired to connect events that are correlated over time, even if they are unrelated or random. Most of the time, the first event did not cause the second event.
· Problems are visible, while benefits are often invisible. When there are problems, there are lots of noisy complaints. When things go well, people rarely notice or speak up.
· Assume all problems are multi-factorial unless data to the contrary is overwhelming. As journalist H.L. Mencken observed, "For every complex problem there is an answer that is clear, simple, and wrong."
· In science, believing doesn't make something true. Numerous conclusions circulating on the Internet lack data or analysis. Nonetheless, many people believe them to be true.
· You can't prove a negative. While science can be overwhelming in one direction, there is rarely a way to prove that there are no possible exceptions.
A hat-tip to Unbiased Science (here) for providing me with a seventh one: “you can’t ethically run placebo-controlled trials on vaccines [and other medical products] that are already known to work and are the standard of care.”
Critical Loss of Manpower and Expertise Can Be Partially Offset By The Growing Role of AI in Regulatory Science
“AI can help bridge some--certainly not all--of the gap created by the loss of expert staff at the FDA. However, this is speculative, not proven...and the progress to date is likely real but almost certainly oversold.”
Thoughts to carry forward into 2026:
The Skeptical Enthusiast’s Guide to AI at FDA explored the incredible potential for AI and supported Commissioner Makary’s AI initiatives. I pointed to a parallel: the transformative impact on government that came about during the 1990’s with the introduction of the Internet.
The column cautioned against expecting short-term gains in productivity or improved regulatory decisions. It also urged step-by-step integration and greater validation of AI tools before the agency relied on them.
Because of layoffs and retirements in 2025, there will be fewer staff in 2026, and the expertise of new hires will be less valuable initially. They will need time to gain experience. In the short term, AI has the potential to partially offset the manpower/expertise gap.
Optimistically, we will start to see some meaningful impact from AI integration in 2026 (see the recent FDA update here), even as it will be several more years before AI is truly transformative.
Inspection and Compliance: Harder and More Important Than Stakeholders, Congress, and the Media Realize
“Probably the single most essential service of FDA is its inspections and compliance work….Going back to the late 1800’s (the “Gilded Age” and pre-FDA), we know what happens if you do not regulate food and drugs.”
Thoughts to carry forward into 2026:
· When it comes to food and medical products, FDA is our nation’s Department of Homeland Security. https://bit.ly/47kUyRe.
· Inspectors and compliance staff...never get sufficient credit for their difficult, behind-the-scenes, and potentially dangerous work (here) overseeing nearly 300,000 FDA-regulated facilities, including farms, food and medical ingredient processing plants, and medical manufacturing facilities (here).
· Even long-time FDA-watchers often don’t understand the strength of the motivation or the size of the profit that is derived from cutting corners and evading FDA regulation. We should have zero tolerance for greed and negligence, which are the most common precursors of unsafe food.
· Recurring issues with baby food and frequent food recalls highlight that much work still needs to be done. The fixes will require resources and commitment that are currently lacking (here).