Liverpool University Professor of Health David Taylor-Robinson discovered his own face selling supplements online.
It was not a parody account. According to reporting from this December, a series of marketing videos used artificial intelligence to impersonate real doctors and public health figures across TikTok, Facebook, X, and YouTube, promoting sketchy products and even inventing fake conditions. Now more than ever, medical authority is a costume, something that can be manufactured on a laptop in seconds.
That should terrify us for reasons that go beyond social media moderation.
Underlying nearly every debate over misinformation, vaccines, climate, or AI, the average person’s alienation from the process of scientific discovery is incredibly damning. The majority of Americans experience science as something done behind closed doors, in a place so far removed from their lives that it might as well be Antarctica. In 2023, 57% of Americans said science has had a mostly positive effect on society, but that share has declined by 16 points since 2019. Science has become socially optional, like a hobby one can opt out of if it makes them uncomfortable.
Here’s my thesis: the most practical anti-misinformation policy for America is expanding access to original research opportunities for middle and high school students.
My personal journey in science research has not been easy. Back in middle school, the Hillview STEM Fair was a major tradition that even other local high schools attended. My friend Aidan Sharp and I investigated how to disinfect contaminated water via boiling, freezing, and other methods—a simple project, but it was the first time I felt the difference between learning science and doing science. Explaining my project to those unfamiliar with the field was the first time I truly understood it, in a way that passive learning in my classes had never enabled me to.
Then the fair disappeared. COVID was the reason given, but it hasn’t returned since. The Menlo Park City School District barely notices because science research is not required for graduation. Even the M-A STEM Fair is significantly underutilized by the entire student body, with only a handful of kids participating each year.
This pattern extends far beyond one district. Nationally, participation in original science research is thin, even when schools offer it. In the High School Longitudinal Study, only 6.3% of students participated in a science competition across the U.S. At the same time, the institutions that make youth research possible are being hollowed out. In May and June 2025, the President’s Fiscal Year (FY) 2026 budget request proposed $3.9 billion for the National Science Foundation (NSF). The American Astronomical Society’s breakdown of the request describes a 56.9% reduction in discretionary spending from the FY 2024 plan, forecasting steep reductions that will affect graduate students, undergraduates, and even K-12 teachers and students.
This stark reality demands that districts make a significant trade-off. Districts should redirect funding away from redundant benchmark testing and standardized curriculum, instead focusing on creating credited, school-day research programs.
Stanford education policy researchers have documented that districts spend an average of $55 per pupil annually on summative testing, excluding time spent on test preparation and low-score interventions. At M-A, with over 2,000 students, that represents over $100,000 in spending annually. That money should instead fund small project grants, basic equipment, and a coordinator who can match students with appropriate mentors and ensure projects are feasible.
Internationally, youth science research initiatives have already made a significant impact. In the United Kingdom, the Nuffield Research Placements resulted in an average student grade increase of 7.5 points (nearly a full letter grade), greater higher-education enrollment, and a 14.3 percentage point higher likelihood of enrolling in a STEM course at college. In Germany, the Jugend forscht initiative resulted in over nine out of ten participants later pursuing a scientific, technological, engineering, mathematical, or medical field of study. And in Singapore, the country with the consistently highest scientific literacy rate in the world, they pride themselves on running science research pipelines for teenagers through programs like the Ministry of Education-backed NUS-MOE and the Junior Researcher initiatives for all students.
It is clear that the deepfake doctor problem is not going away in 2026. The costumes will just get better and cheaper. In a world where expertise can be convincingly faked, trust cannot be rebuilt solely through online moderation and policing misinformation. Trust must be earned earlier, through direct experience with how scientific claims are tested, challenged, and often proven wrong.
