The AI Slop Scandal Around the MAHA Report Is Getting Worse
It came to light this week that a new government report from the "Make America Healthy Again" Commission led by Robert F Kennedy Jr. contained botched citations for scientific papers that didn't exist. This is almost certainly a sign that some form of generative AI was involved to draft a very consequential piece of medical agenda-setting, coming out of the US's top health agency, the Department of Health and Human Services.
Now, some additional reporting suggests that the paper's flaws go even deeper — yes, even deeper than allegedly relying on a technology known for making stuff up and then being surprised that it made stuff up.
But first, let's highlight how the White House finally decided to respond to the criticism of the report, which has been "very poorly and not convincingly at all."
On Thursday, the White House said that it would fix the errors in the government report — and it did, releasing a new version with corrected citations. But press secretary Karoline Leavitt also took the opportunity to construe the affair as the press getting worked up about a few errant typos.
"I understand there was some formatting issues with the MAHA report that are being addressed and the report will be updated," Leavitt told reporters during a press briefing, as quoted by the Associated Press. "But it does not negate the substance of the report."
"Minor citation and formatting errors have been corrected," HHS spokesman Andrew Nixon told the AP in a statement.
Led by noted anti-vaxxer and all-around crackpot RKF Jr., the "MAHA Report" purports to be a tell-all on why Americans, and especially children, are so unhealthy. Both Leavitt and Nixon described the report as "transformative."
That's a questionable claim. As NOTUS first reported on Thursday, several of the studies cited in the report do not exist at all, including one called "Overprescribing of Oral Corticosteroids for Children With Asthma," which was used to argue that doctors are giving kids too much medicine. This "study" has never been referenced anywhere outside the MAHA report. Lawyers have been sanctioned for similar behavior in court.
It gets dumber. The Washington Post found that 37 of the paper's 522 footnotes are inexplicably repeated multiple times. Some of the citations also include an "oaicite" appended to the URLs, which refers to OpenAI, the creators of ChatGPT. This is a "definitive sign" that the research was gathered using an AI, WaPo concluded.
And the flaws go beyond bogus citations, or "minor" perils of "formatting," in Leavitt's parlance. As experts told the NYT, some of the papers that were correctly cited were still inaccurately summarized — if in fact they weren't being deliberately misconstrued. The report argued, in one case, that a 40-fold increase in bipolar disorder and ADHD diagnoses in children between 1994 to 2003 was propelled by loosened criteria in a fifth edition of a guide used by psychiatrists, per the NYT. But that fifth edition, it turns out, didn't come out until 2013. And that "40-fold increase" the report touted appears to come from a 2007 study which makes zero mention of an uptick in ADHD.
Even if you could somehow excuse using an AI chatbot to help speed up composing what's supposed to be the cynosure of the US's public health policy going forward, the sheer levels of sloppiness on display can only leave you to conclude that either RFK and his lackeys have no idea what they're doing and have no business writing a serious scientific document, or that they're trying to pull the wool over the eyes of the American public, cynical enough to ask an AI model to conjure up studies to fit whatever narrative they're peddling.
"This is not an evidence-based report, and for all practical purposes, it should be junked at this point," Georges C. Benjamin, executive director of the American Public Health Association, told WaPo. "It cannot be used for any policymaking. It cannot even be used for any serious discussion, because you can't believe what's in it."
More on AI: This Sleazy GLP-1 Prescription Site Is Using Deepfaked "Before-and-After" Photos of Fake Patients, and Running Ads Showing AI-Generated Ozempic Boxes
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Boston Globe
25 minutes ago
- Boston Globe
Trump's book bans hurt the chances for reparations
'We are in a moment of anti-Blackness on steroids, and we refuse to be silent,' Pressley said earlier this month during a press conference at the US Capitol. 'We will not back down in our pursuit of racial justice,' she added. 'The antidote to anti-Blackness is to be pro-Black, and we will do it unapologetically. The United States government owes us a debt, and we need reparations now.' A large majority of Black Americans agree with Pressley. Nearly 3 in 4 Black adults support reparations, according to a 2024 Advertisement But the percentage of Americans of other races and ethnicities who back the idea is low. Less than half of Hispanic (47 percent) and Asian American (45 percent) respondents are in favor of reparations. And only about a third (34 percent) of white adults surveyed back the idea. Only 36 percent of Americans overall back the idea, according to The reasons for this vary. Some of it may be rooted in prejudice and bias. After all, Japanese Americans received Advertisement But some of the opposition to reparations is rooted in ignorance. As communities prepare to commemorate the ending of slavery later this month on Juneteenth, the majority of Americans finish high school knowing very little about just how atrocious slavery was. Only 8 percent of high school seniors were able to identify slavery as the central cause of the Civil War, according to a 2018 Southern Poverty Law Center And in 2017, There are long-term consequences for this knowledge gap. Just 1 in 4 adults (24 percent) strongly agree that the legacies of slavery affect the position of Black people in American society today, according to the Princeton survey. And America's ignorance about slavery is likely to become more widespread given that support for book bans has reached the federal level. In an executive order aimed at preventing students from reading books that introduce ideas about privilege and oppression and their relationship to race, President Trump accused schools that teach students books like Advertisement 'Such an environment operates as an echo chamber, in which students are forced to accept these ideologies without question or critical examination,' he It's understandable why Trump, who made white grievance a foundational part of his presidential campaign, believes his effort to silence authors is popular. His return to the White House is largely viewed as confirmation of many Americans' rightward shift — even on matters of race — since the summer of 2020, when people filled the streets across the country to protest anti-Black racism after the police killing of George Floyd. Most Americans But Trump is misguided. Americans may not be in favor of what they consider preferential treatment based on race. But they are not fans of banning books — including those that aim to make a case for the need for that preferential treatment. Two-thirds of Americans oppose efforts to restrict books in public schools, according to a 2024 Knight Foundation While former vice president Kamala Harris was unsuccessful in her attempt to keep Trump from returning to the White House, she seemed much more in line with where most Americans are when it comes to learning about this country's history. Advertisement 'We just need to speak truth about history. In spite of the fact that some people try and erase history and try and teach our children otherwise,' Harris 'We need to speak truth about it in a way that is about driving solutions,' added Harris, who as a senator cosponsored the bill that Pressley recently reintroduced. In a country where
Boston Globe
25 minutes ago
- Boston Globe
Can elite universities remain global?
One reason the schools have arguably been caught off guard is less so: They have misread the nation. Advertisement Large and influential swaths of America The globalization of America's universities began decades ago. When Richard Levin assumed the presidency of Yale in 1993, in his inaugural address he stated that 'as we enter the 21st century, we must aspire to educate leaders for the whole world. … We must focus even more on global issues … if we are to be a world university.' For Levin, the mission was clear. In the early 18th century, Yale's mandate was to educate leaders and citizens for the region. 'By the mid-nineteenth century, our compass had become the whole nation,' he said. Now the work would be global. And Yale was far from alone in such ambitions. Advertisement From the establishment of campuses overseas to the creation of research centers and collaborations worldwide and the embrace of international students, many universities have changed dramatically in the years since Levin made those remarks. Today, For years, this embrace of international students was largely seen as in accord with the national interest. Despite pockets of protests, globalization on campus was treated as inevitable — and desirable — in many quarters. Meanwhile, the federal government maintained expansive investment in these universities — to the tune of The assumptions driving the internationalization of America's universities, however, have now changed. Many people no longer believe globalization is good for America. That change is most obvious in the MAGA movement. But the anti-Israel protests tinged with anti-American, anti-Western, and anti-capitalist messages that some international students have helped lead have alienated other Americans as well. Advertisement As a result, many universities have been caught out over their skis. Some now see a heavy international student presence less as a virtue and more as something suspect when it comes to university leaders' motives and wisdom. On university campuses, many faculty would undoubtedly disagree and argue that globalization is still a positive force for America. But with seats scarce at exclusive universities, filling them with international students is seen through a zero-sum lens. Universities may now face a decision. Do they want internationalization or federal support? Both may no longer be an option. The compact between universities and the federal government can only continue if the work of the university is seen as being in lockstep with the national interest. This isn't to say this is the end of global universities. Or of research universities. Minerva University, where I serve on the board of trustees, is unabashedly global. Eighty-five percent of students hail from countries other than the United States. All students live outside the United States for three of their four years. But Minerva doesn't take any federal money, nor is its model built around research. On the other hand, the Highly selective universities may choose to fight to retain federal support and remain global in the hopes that they can weather the next few years. Advertisement But with alternative models and the ascendance of skepticism around the merits of globalization, it seems less clear if this will be a viable strategy.
Yahoo
an hour ago
- Yahoo
Here's how Uber's product chief uses AI at work — and one tool he's going to use next
Uber's product chief said he uses AI to summarize lengthy reports and do research before launches. Sachin Kansal uses ChatGPT and Gemini to understand Uber's performance in overseas markets. He plans to add Google's NotebookLM to his AI suite. Uber's chief product officer has one AI tool on his to-do list. In an episode of "Lenny's Podcast" released on Sunday, Uber's product chief, Sachin Kansal, shared two ways he is using AI for his everyday tasks at the ride-hailing giant and how he plans to add NotebookLM to his AI suite. Kansal joined Uber eight years ago as its director of product management after working at cybersecurity and taxi startups. He became Uber's product chief last year. Kansal said he uses OpenAI's ChatGPT and Google's Gemini to summarize long reports. "Some of these reports, they're 50 to 100 pages long," he said. "I will never have the time to read them." He said he uses the chatbots to acquaint himself with what's happening and how riders are feeling in Uber's various markets, such as South Africa, Brazil, and Korea. The CPO said his second use case is treating AI like a research assistant, because some large language models now offer a deep research feature. Kansal gave a recent example of when his team was thinking about a new driver feature. He asked ChatGPT's deep research mode about what drivers may think of the add-on. "It's an amazing research assistant and it's absolutely a starting point for a brainstorm with my team with some really, really good ideas," the CPO said. In April, Uber's CEO, Dara Khosrowshahi, said that not enough of his 30,000-odd employees are using AI. He said learning to work with AI agents to code is "going to be an absolute necessity at Uber within a year." Uber did not immediately respond to a request for comment from Business Insider. On the podcast, Kansal also highlighted NotebookLM, Google Lab's research and note-taking tool, which is especially helpful for interacting with documents. He said he doesn't use the product yet, but wants to. "I know a lot of people who have started using it, and that is the next thing that I'm going to use," he said. "Just to be able to build an audio podcast based on a bunch of information that you can consume. I think that's awesome," he added. Kansal was referring to the "Audio Overview" feature, which summarizes uploaded content in the form of two AIs having a voice discussion. NotebookLM was launched in mid-2023 and has quickly become a must-have tool for researchers and AI enthusiasts. Andrej Karpathy, Tesla's former director of AI and OpenAI cofounder, is among those who have praised the tool and its podcast feature. "It's possible that NotebookLM podcast episode generation is touching on a whole new territory of highly compelling LLM product formats," he said in a September post on X. "Feels reminiscent of ChatGPT. Maybe I'm overreacting." Read the original article on Business Insider Error in retrieving data Sign in to access your portfolio Error in retrieving data Error in retrieving data Error in retrieving data Error in retrieving data
