Harvard’s Top Example of Federally Funded Research Is Actually Surprisingly Weak

Overly hyped cancer research — by 12 doctors in China

An anti-Israel protest featuring Harvard faculty near Harvard Medical School, Boston, January 6, 2025.

What valuable project will grind to a halt because of the federal government’s freeze on $3.2 billion worth of research funding to Harvard?

You’d think that with all that sponsored research to choose from, Harvard’s 16 outside lawyers, in combination with the university’s vast staff of in-house attorneys, public-relations and government-relations staff, research administrators, and deans would have an easy time finding a super-compelling example.

Someone high up at Harvard seems to believe, reasonably enough, that it will help Harvard get the $3.2 billion flowing if the university can make the court and the public believe that President Trump and Secretary of Health and Human Services Robert Kennedy Jr., are hurting innocent cancer victims. In the federal complaint Harvard filed Monday afternoon in U.S. District Court for Massachusetts, the word “cancer” appears 13 times, and “oncology” twice.

“Federally funded research at Harvard has led to life-altering advancements in improving cancer diagnoses,” the complaint says.

What’s the evidence?

That sentence of the complaint includes a footnote to a September 4, 2024, article by Ekaterina Pesheva, who is “senior director, science communications & media relations” at Harvard Medical School. The headline is “A New Artificial Intelligence Tool for Cancer” and the subheadline, with a healthy dose of that famously characteristic Harvard humility, is “ChatGPT-like AI model can diagnose cancer, guide treatment choice, predict survival across multiple cancer types.”

Life-altering? Not so fast. Even the Harvard article, which is basically a press release, eventually makes clear that this tool hasn’t altered anyone’s life quite yet. The article quotes a Harvard assistant professor of biomedical informatics Kun-Hsing Yu, who says, “If validated further and deployed widely, our approach, and approaches similar to ours, could identify early on cancer patients who may benefit from experimental treatments targeting certain molecular variations, a capability that is not uniformly available across the world.” Note the speculative “if” and “could,” which are absent from the more definitive, past tense claims in Harvard’s legal complaint of “has led to life-altering advancements.”

You don’t even have to take my word or Professor Kun-Hsing Yu’s word for it. The director of the National Institutes of Health during the Biden administration, Dr. Monica Bertagnolli, a breast cancer survivor, published an October 3, 2024 blog post reporting on the work but also cautioning, “there’s much more work ahead before an AI model like this could be used in the clinic.” She spoke of the need for “continued development and validation.”

I emailed Professor Kun-Hsing Yu and his administrator, Faith McDonald, this morning, to ask, “Do you think it's accurate to describe your advance as ‘life altering’? Has anyone’s life been altered by it yet, actually? Or is that only a hope for the future?” They haven’t gotten back to me yet with an answer.

So, in Harvard’s own cherrypicked example of a particularly important piece of federally funded cancer research, Harvard’s lawyers have overstated the significance of the achievement.

If that seems a subtle, arcane, or pedantic point, just wait: it gets even juicier.

The Harvard article and the NIH blog post both are based on an article that appeared on September 4, 2024 in the scientific journal Nature. Headlined, “A pathology foundation model for cancer diagnosis and prognosis prediction,” that article carried the bylines of 29 authors. Of those, 12 were based in China. They are Wei Yuan, Jiayu Zhang & Jing Zhang of the College of Biomedical Engineering, Sichuan University, Chengdu, China; Jietian Jin of the Department of Pathology, Sun Yat-sen University Cancer Center, Guangzhou, China; Hongping Tang of the Department of Pathology, Shenzhen Maternity & Child Healthcare Hospital, Shenzhen, China; Kanran Wang of the Department of Radiation Oncology, Chongqing University Cancer Hospital, Chongqing, China; Yu Li of the Department of Pathology, Chongqing University Cancer Hospital, Chongqing, China; Fang Wang of the Department of Pathology, The Affiliated Yantai Yuhuangding Hospital of Qingdao University, Yantai, China; Yulong Peng of the Department of Pathology, The First Affiliated Hospital of Jinan University, Guangzhou, China; Junyou Zhu of Department of Burn, The First Affiliated Hospital, Sun Yat-sen University, Guangzhou, China; and Jun Zhang & Xiao Han of Tencent AI Lab, Shenzhen, China. (Harvard’s extensive China ties more generally, not having to do with this particular project, are the topic of a paper out this week from Isaac Stone Fish.)

I emailed Professor Kun-Hsing Yu and his administrator, Faith McDonald, this morning, to ask, “Was the U.S. taxpayer funding via Harvard used to support the China-based research?” They haven’t gotten back to me yet with an answer.

Yet the news of the extensive Chinese involvement in the Harvard cancer research undercuts an argument in Harvard’s court complaint, which says “each suspended research project hinders the cultivation of the scientific talent that drives the United States’ global competitiveness in research and development,” or, as Harvard President Garber puts it, quoted in the complaint, the funding freeze risks “the economic security and vitality of our nation.” It’s harder to make the case that Harvard’s federally sponsored research helps “the United States’ global competitiveness” when 41 percent of the doctors on the research project are based in China.

What’s more, the ethics declarations and acknowledgments on the Nature article, along with those in the Harvard Medical School article, indicate that the research work was supported not only by federal grants but also by Amazon, Microsoft, Nvidia, Google, the American Cancer Society, and the South Korean government. Two authors are employees of Tencent AI Lab, Tencent being a China-based tech company with a $561 billion market capitalization (roughly ten times Harvard’s endowment). Harvard also described Kun-Hsing Yu as a consultant to Takeda, a giant Japanese pharmaceutical company.

I emailed Professor Kun-Hsing Yu and his administrator, Faith McDonald, to ask whether any of those funding sources might be possibilities to continue to support the research work if U.S. federal funding is eliminated or frozen. He hasn’t yet gotten back to me with an answer.

Had I thought of it, I might have also asked why he doesn’t just transfer the federal grant out of Harvard and to another researcher’s collaborating institution that isn’t in trouble with Trump, such as Pennsylvania State University in Pennsylvania or Cedars-Sinai Medical Center in Los Angeles.

The point here isn’t to pick on Professor Kun-Hsing Yu or his artificial-intelligence powered pathology slide image analysis program. Anything that produces more accurate and less expensive cancer diagnoses will certainly be greatly appreciated by patients and their loved ones, myself included. It may be that as a legal matter, Harvard is correct that the Trump administration failed to follow the proper procedures before freezing Harvard’s grant money. There are lots of other addressable questions for another day on whether this funding freeze and the Trump administration’s approach to Harvard (or Harvard’s approach to the issues the federal government is complaining about) are optimal. I didn’t select this particular project. Harvard chose to highlight it in its court case.

But if, of all the $3.2 billion in federally funded projects, the best example Harvard can come up with is one that the university misleadingly oversells the accomplishments of, is nearly half-based in China, and has a raft of deep-pocketed possible alternative funders standing by to pitch in for, then it may well be that Harvard’s case for a continued claim on vast flows of federal research funding is substantially weaker than the university’s leaders and its lawyers would have the court and the public believe.