The advent of fact-checker journalism may be wearing out its welcome. Perhaps the increasing politicization of American life is a contributor to the downward spiral of the fact-checking profession that is primarily run by politically engaged reporters, not expert specialists in the subjects they assess by any sense of the imagination. Not that any one group of experts should have the authority over the truth either. Self-appointed media gatekeepers are a ticking time bomb of political censorship, waiting to be unleashed when the temptations are too great and the necessity for impartiality is even greater. With White House Press Secretary Jen Psaki calling for collusion between social media companies and the government to censor “misinformation”, this threat seems to be as close as ever.
Although fact checkers purport to be independent guardians of accountability, recent events have exposed them as mere enforcers of fashionable political positions. This brings us to a relatively new, but powerful company known as NewsGuard, which claims a partnership with Microsoft and gleaming spotlights in major outlets. Its staff and board boast powerful connections to the government, finance, and the media. According to an Op-ed in Politico written by NewsGuards’ CEO, rather than simply being a fact-checking company that can only debunk stories after they go viral, NewsGuard rates entire websites’ trustworthiness. This new strategy is aimed at discrediting the very source that alleged misinformation or disinformation may come from. NewsGuard publishes lengthy “nutritional labels,” rating websites on various criteria of journalistic importance and outlining its reasons for giving certain ratings. Perhaps one day, these ratings may be used to filter out certain websites, which is what NewsGuard’s CEO alludes to by citing the great political scientist Francis Fukuyama’s article in Foreign Affairs.
In fact, the company made the following tweet on July 17, 2021, essentially siding with Psaki’s call for a government-media partnership to censor internet content.
Figure I: NewsGuard tweet praising Psaki’s statement
Phrases such as “working collaboratively” and “our door is open” reveal that NewsGuard envisions itself as having a prominent role in any collusive arrangement that arises from the White House’s recent statements. Their posturing for influence, however, obscures a more basic question: why should we trust a self-appointed fact checker? Indeed, who fact checks the fact checkers?
After receiving a recent request for comments on a “fact-check” article by NewsGuard regarding AIER and the Great Barrington Declaration, we decided to investigate the rise of the fact-checking phenomenon itself, including this strange new company’s own performance in evaluating the content of other websites.
We soon discovered that NewsGuard falls far short of the very same criteria for accuracy and transparency that it claims to apply to other websites. Most of the company’s fact checkers lack basic qualifications in the scientific and social-scientific fields that they purport to arbitrate. NewsGuard’s own track record of commentary – particularly on the Covid-19 pandemic – reveals a pattern of unreliable and misleading claims that required subsequent corrections, and analysis that regularly conflates fact with opinion journalism in rendering a judgement on a website’s content. Furthermore, the company’s own practices fall far short of the transparency and disclosure standards it regularly applies to other websites.
How Accurate are Self-Appointed Fact Checkers?
The stated premise of fact-checking is to correct errors in claims published by other websites and outlets. But what happens when the organization doing the fact-checking has its own facts wrong?
A revealing example may be found in NewsGuard’s treatment of the “lab leak” hypothesis for Covid-19’s origins. Media coverage of the lab leak theory – which posits that the pandemic originated through the accidental infection of workers at the Wuhan Institute of Virology who were studying coronaviruses in bat populations – has changed dramatically in recent months after a closer examination of evidence led several scientists to lend it credence. Political figures including President Joe Biden and White House medical adviser Anthony Fauci now consider the lab leak theory “plausible” and have called for an investigation of the Wuhan facility.
For over a year prior to these recent developments however, NewsGuard aggressively “fact checked” and penalized other websites for even raising the possibility of a lab leak. Some of the most aggressive attacks came from John Gregory, NewsGuard’s “Deputy Editor for Health” policy and also the primary correspondent in AIER’s exchanges with the company.
In March 2020, Gregory sent a separate inquiry to another website charging it with promoting “unfounded conspiracy theories about the virus’ origins,” and specifically what is known as the lab leak hypothesis. “There is no evidence that the Wuhan Institute of Virology was the source of the outbreak, and genomic evidence has found that the virus is ‘96% percent identical at the whole-genome level to a bat coronavirus,’” continued Gregory’s email, which also suggested that NewsGuard would be flagging and downgrading the website for publishing “misinformation” about this subject. While the website in question – an “alternative medicine” blog – promoted other fringe claims that warranted scrutiny and legitimate corrections, Gregory’s focus on specifically discrediting the lab leak thesis suggests he was injecting his own political opinions into the fact-checking exercise.
In another example from February 2020, Gregory announced by tweet that he had contacted a medical news website after it ran a headline suggesting “Coronavirus may have leaked from China’s highest biosafety lab.” Gregory demanded a “formal correction or retraction” of the headline. After we contacted Gregory by email to question him about the practice of penalizing websites for discussing the lab leak hypothesis, he responded, “My February 2020 tweet was also accurate,” asserting that his position was justified on the grounds that “[t]he lab leak theory remains unsubstantiated and under investigation.” NewsGuard has taken a similar line in its reassessment of ratings on some 225 websites where it docked the source for mentioning the lab leak hypothesis, stating that “while not substantiated, [the lab leak theory]is also, as of now, not provably false.”
Despite Gregory’s tendentious phrasing, he nonetheless quietly deleted the tweet in the days following AIER’s inquiry. The company also found itself in deeper trouble over its previous mischaracterizations of the lab leak hypothesis as a “conspiracy theory” – a common refrain in Gregory’s articles on the subject. In late June 2021, NewsGuard had to issue at least 21 separate corrections to ratings where it docked websites for reporting on the lab leak theory. According to a statement that the company sent to AIER:
“NewsGuard either mischaracterized the sites’ claims about the lab leak theory, referred to the lab leak as a “conspiracy theory,” or wrongly grouped together unproven claims about the lab leak with the separate, false claim that the COVID-19 virus was man-made without explaining that one claim was unsubstantiated, and the other was false. NewsGuard apologizes for these errors. We have made the appropriate correction on each of the 21 labels.”
NewsGuard has not published the full results of its audit, or a list of the corrections it made, thereby precluding independent verification of whether its corrections were sufficient to rectify the Covid misinformation it had previously published. The company did not respond to AIER’s request for this information.
While these corrections are a welcome development from the company, they also reveal a deeper underlying hubris that characterizes NewsGuard’s general approach to vetting Covid-19 content. Gregory and other fact checkers at the site appear to have concluded as early as February and March 2020 that the lab leak hypothesis was nothing more than a fringe “conspiracy theory,” and began using this descriptor to attack and downgrade almost any website that subsequently raised the very possibility that it was investigation-worthy. This hubris persisted until a reassessment of the claim led mainstream scientists, as well as political figures such as Biden and Fauci, to deem the hypothesis “plausible” and call for a comprehensive investigation into the Wuhan Institute of Virology. Gregory and his colleagues appear to have simply decided that their own premature dismissal of the lab leak hypothesis equated to “fact” and proceeded to penalize other sites not for factual errors, but rather for diverging from NewsGuard’s own editorial position on the same subject. When this position turned out to be mistaken, NewsGuard pivoted to remove the errors – albeit in non-transparent ways that downplay the significance or pervasiveness of their mistake.
AIER’s own experience with NewsGuard revealed a similar pattern of carelessness and misrepresentation by Gregory and other writers for the company. Gregory contacted us on behalf of NewsGuard in early June 2021 requesting comments on several articles relating to Covid-19 pandemic policy and the Great Barrington Declaration. AIER’s Phil Magness obliged the request by offering to answer his questions in good faith, but quickly discovered that they carried heavy political biases arising from Gregory’s own personal beliefs about Covid-19, healthcare policy, American politics, and related subjects.
In one such example, Gregory asked a prejudicial question that attempted to implicate AIER with showing partisan political biases in our publications:
“We also note that AIER.org refers to itself as nonpartisan. Why then do its articles routinely criticize Democrats, such as a June 2021 article that stated, “Biden has never provided any evidence that he is more trustworthy on corruption than any other career Washington politician,” or another recent article that said Biden, Fauci, and House Speaker Nancy Pelosi “do not care if their policies destroy the nuclear family, the educational system, and moral and religious value systems, the very pillars of civil society?”
Gregory’s question, however, selectively cherry-picked only two articles on our site where we criticized Democratic politicians. It made no mention of the many examples where AIER has similarly criticized Republicans such as former president Donald Trump over excessive spending, tariffs, support for lockdowns, and illiberal political posturing. As Magness responded, “[T]he focus of [AIER’s] criticism is also shaped by who happens to be in political power at the time.” Gregory’s line of questioning about Democrats appears to have intentionally ignored “numerous instances where we have similarly criticized Republicans” when they were in power.
Gregory’s questions displayed a similar pattern of conflating normative policy positions taken by individual authors on AIER’s website – essentially opinion articles, and all properly identified as such – for positive or empirical claims, which could then be “fact checked.” As a result, his questions treated prescriptive policy opinions that diverged from his own viewpoints as if they were “scientific claims” – even when the normative nature of the argument was explicitly stated up front (in one example, the authors of a “fact checked” argument prefaced their opinions on how vaccines should be prioritized as their own beliefs, including identifying it as “our views on this” subject). Since AIER publishes a diversity of positions, including arguments that diverge from the viewpoints of AIER’s full time research faculty, it is fundamentally inaccurate to portray individual opinion essays as “AIER.org’s claim,” as NewsGuard does. Indeed, Magness has written at length in favor of expanding vaccination access despite FDA and CDC regulatory obstacles, taking the exact opposite position of the viewpoint that Gregory attributes to the organization. When Magness replied to Gregory by calling attention to the difference between normative and positive arguments as well as the editorial diversity of external contributors to our daily publications, he ignored the distinction. Gregory’s subsequent article proceeded to blur the two together, erroneously depicting normative points of disagreement as positive scientific claims that could then be “fact checked” against his own position.
Even more problematic was NewsGuard’s portrayal of the Great Barrington Declaration (GBD), signed at AIER in October 2020. Gregory’s synopsis of the GBD contained numerous false and misleading claims that were brought to the attention of his company almost immediately after their publication.
Repeating a charge from another website, Gregory wrote that “none of the three [GBD authors] had published peer-reviewed research about the COVID-19 pandemic at the time they authored the declaration.” This claim is false. GBD co-author Jay Bhattacharya was part of a team of scientists from Stanford University that conducted one of the first wide-scale seroprevalence studies of Covid-19 at the outset of the pandemic. Their results appeared in the Journal of the American Medical Association in May 2020. When contacted by AIER about this error in his article, Gregory conceded that the claim “will require a correction on our part,” though he appended it with a snide denigration of Bhattacharya for being “listed as the seventh author” on the study (Bhattacharya was in fact a principal co-author but was listed last, as per a convention with how some medical journal articles identify senior ranked investigators. Bhattacharya was also a primary media contact about his study’s findings at the time of its release).
NewsGuard’s depiction of the GBD contained other clear misrepresentations of its contents and positions. For example, Gregory wrote that the GBD “argued that restrictions meant to reduce the spread of the COVID-19 virus, such as face masks…should be eliminated for people considered to be at lower risk of severe illness and death from COVID-19.” The text of the GBD makes no mention of face mask policy though – only lockdowns and similar restrictions on schools and businesses. NewsGuard did not respond to multiple requests from AIER to correct this erroneous characterization.
Gregory’s article also displayed a clear pattern of relying upon dubious and unqualified secondary sources to evaluate the scientific merits of the GBD. In an email to AIER, NewsGuard co-CEO Steven Brill stated “when we make judgments about health care sites…we rely on – and quote — sources who are the experts.” This is not the case with their assessment of the GBD.
Rather than quoting scientific experts, NewsGuard’s review of the GBD relies primarily on a statement by former UK Health Secretary Matt Hancock – a politician who has no formal scientific or medical training. In a passage quoted by Gregory, Hancock stated that “the Great Barrington declaration is underpinned by two central claims and both are emphatically false. First, it says that if enough people get covid, we will reach herd immunity. That is not true…we should have no confidence that we would ever reach herd immunity to covid, even if everyone caught it.”
Hancock’s statement, however, is at direct odds with mainstream science on immunology. The World Health Organization specifically defines herd immunity as the combined total of immunity acquired by vaccination and by natural infection and recovery. Although it differs from the GBD authors on how to most effectively reach this point, the WHO does not dispute the existence or attainment of herd immunity itself. Even among pro-lockdown sources, herd immunity is seen as a reachable goal and a primary aim of mass vaccination, contrary to Hancock’s claims. Dr. Anthony Fauci, for example, has frequently stated that he anticipates reaching herd immunity after between 70 and 85% of the U.S. population is vaccinated or recovered from Covid-19. Most other public health scientists agree that herd immunity is attainable through a combination of vaccination and natural immunity, even though some of them differ from the Great Barrington Declaration authors about the role of specific policy interventions in reaching this point.
Hancock’s statement, cited as authoritative by NewsGuard, further contended, “The second central claim [of the GBD]is that we can segregate the old and vulnerable on our way to herd immunity. That is simply not possible.” This is not a scientific statement, but rather Hancock’s own political opinion. A detailed plan arguing for the feasibility of focused protection measures was published by the GBD authors to accompany the Declaration itself. More importantly, the scientific literature on Covid-19 mitigation documents clear evidence that the success (or failure) of a country to “shield” its nursing homes through a focused protection strategy is a primary factor in its overall mortality rate. A study by John P.A. Ioannidis in the journal BMJ-Global Health compared the nursing home shielding ratios of several countries, concluding that they “varied markedly in the extent to which they protected high-risk groups.” Contrary to Hancock’s political claims, Ioannidis concluded: “Both effective precision shielding and detrimental inverse protection can happen in real-life circumstances. COVID-19 interventions should seek to achieve maximal precision shielding.”
When asked by AIER about their continued reliance on Hancock as a source despite the scientific misinformation contained in his assessment of herd immunity as well as his overall lack of scientific qualifications, Gregory responded that it “was and is relevant to explaining the views of those who criticized the Declaration.” Neither Gregory nor NewsGuard responded to follow-up questions about how they reconciled this position, the political nature of Hancock’s comments, or Hancock’s lack of scientific credentials with Brill’s assertion that they “rely on – and quote — sources who are the experts” in the subject matters they evaluate.
In addition to using non-expert assessments such as Hancock to attack the GBD, Gregory included links to further readings about the Declaration from an extremely dubious source: 9/11 Truther and conspiracy theory blogger Nafeez Ahmed of the Byline Times website. Between October 2020 and the present, Ahmed has promoted a flurry of increasingly bizarre conspiracy theories about the GBD, including a false allegation that it was secretly financed by libertarian billionaire Charles Koch in apparent coordination with the British Ministry of Defence and – strangest of all – the proprietor of a resort hotel located in Wales. Far from being based in factual journalism, Ahmed’s theories display clear signs of its author’s personal paranoias and political biases.
When asked about NewsGuard’s promotion of links to Ahmed’s blog, Gregory backtracked by noting that “[NewsGuard] sometimes include articles that we come across while researching a website but are not referenced within the label itself.” He stated that Ahmed’s claims were not used to calculate NewsGuard’s ratings and were only included to provide a “history” of the GBD. It did not appear to concern Gregory that Ahmed’s “history” was an unreliable conspiracy theory of his own imagination.
As of this writing, NewsGuard has not altered or removed the links to Ahmed’s allegations despite its promotion of documented falsehoods about the origins and funding of the GBD. Even more astounding, NewsGuard currently rates Ahmed’s blog with a score of 82.5/100, giving it full credit for “gathering and presenting information responsibly.” This pattern evinces a clear double standard in which NewsGuard promotes sources that do not appear to meet their own published minimum standards for reliability and uses them to denigrate the credibility of AIER and the GBD.
In sharp contrast to the generally disparaging approach he took to covering the GBD, Gregory holds other websites that attack the GBD in high esteem. In one example, Gregory extended a score of 87.5/100 to CovidFAQ.co, a website set up by a group of pro-lockdown activists in the United Kingdom. CovidFAQ is a joint project of conservative member of Parliament Neil O’Brien, “neoliberal” activist Sam Bowman, and academic Stuart Ritchie. Pro-lockdown UK political strategist Dominic Cummings recently referenced their work as part of a “decentralised” political campaign to discredit the anti-lockdown movement and the GBD, which he proposed while serving as an advisor to Prime Minister Boris Johnson. In his article for NewsGuard, Gregory credits CovidFAQ for “not repeatedly publish[ing]false content.” The website’s track record is at clear odds with Gregory’s assessment.
In January 2021, CovidFAQ published a lengthy attack on the GBD that contained multiple errors and misrepresentations of the Declaration’s contents. In one example, the authors of CovidFAQ claimed, “The authors of the Great Barrington Declaration have never given an answer” to how they would implement a focused protection strategy in place of lockdowns. In reality, the GBD website contains a detailed 1,800 word plan for implementing focused protection. When AIER’s Phil Magness alerted CovidFAQ co-owners Stuart Ritchie and Sam Bowman to this error in January 2021, the website’s editors modified the text to read, “The authors of the Great Barrington Declaration have never given anything approaching an adequate answer” to how they would implement focused protection (emphasis added). Rather than a factual correction, CovidFAQ’s change inserted their own editorial commentary expressing disagreement with the GBD’s published focus protection strategy as a way of disguising CovidFAQ’s earlier misrepresentations of the document. Gregory’s NewsGuard rating of the CovidFAQ website specifically linked to CovidFAQ’s deceptive edit about the GBD, and described it as having met “NewsGuard’s standards for regularly issuing corrections.”
Figure II: Misleading edit on the CovidFAQ website, classified as an appropriate “correction” by NewsGuard
Typifying other slanted media coverage of the GBD, NewsGuard concluded its assessment by repeating a false story from October 2020, claiming that the GBD’s signature list contained fake names such as “Dr. Johnny Bananas” to inflate its signature count. This story misrepresents the products of an intentional hoax by pro-lockdown journalists including the aforementioned Nafeez Ahmed to flood the website with false signatures. In reality, “Dr. Johnny Bananas” and similar hoax submissions were removed from the GBD website within a few hours of their discovery. An audit of signatures conducted by AIER found that false names amounted to only 0.1% of total signatures on the GBD prior to their removal, with the largest cluster of false names deriving from Ahmed’s hoax campaign on October 9th. NewsGuard did not include any of this context in its article, nor did Gregory permit AIER an opportunity to comment on the misinformation contained in its account of the false signatures.
To briefly summarize, NewsGuard’s coverage of Covid-19 policy and the GBD in particular suffers from a recurring pattern of frequent errors that warrant correction, reliance on fact checkers and other figures who lack qualifications to make scientific assessments, biased depictions designed to disparage or undermine the scientific credibility of the petition, and the promotion of false information from dubious secondary sources, rather than the “scientific experts” it claims to use. In sharp contrast, NewsGuard writers such as John Gregory take a friendly and non-scrutinizing stance toward pro-lockdown opponents of the GBD such as CovidFAQ website – even when they spread factual misinformation about the GBD’s contents and engage in duplicitous editing under the guise of issuing a “correction.” The self-appointed fact checker, it would appear, suffers from a biased and deficient internal fact-checking process for its own work on Covid-19.
Is NewsGuard Qualified to Police Facts?
Although there should be no barriers to entry when it comes to how qualified one should be to comment on an issue, it would help, especially for credibility, to have some sort of relevant background in the subjects being evaluated. Degrees in medicine and other hard sciences are always great when it comes to Covid-related topics. Advanced qualifications in social sciences like economics, policy, government, and law are similarly pertinent, given the pandemic’s wide-reaching public policy consequences. Several of these fields come with advanced training in the gathering of evidence, statistical analysis, and applying the scientific method – all necessary tools to evaluate the performance of policies such as mask mandates, lockdowns, programs to encourage and distribute vaccination, and the sort.
By contrast, NewsGuard’s staff primarily evaluates scientific claims by appealing to the authority of public figures who they designate as “experts” on the subject in question. Their approach generally avoids direct examination of the evidence surrounding contested claims, and instead cherry-picks a figure to treat as an authoritative final word. As their liberal use of Hancock to evaluate the GBD illustrated, many of their preferred authorities are political officeholders rather than persons trained in scientific or social-scientific methods.
By selectively curating cherry-picked political authorities rather than evaluating evidence directly, NewsGuard’s approach to fact-checking effectively sidesteps the scientific method. This strategy is rendered even more problematic by the general lack of scientific expertise within NewsGuard’s team of writers.
We examined the educational credentials, including the highest degree listed, for 28 publicly identified staff members on NewsGuard’s website. The company’s staff page reveals shockingly little expertise in either the hard sciences such as medicine or social sciences such as public policy, economics, and related fields. 12 of the 28 listed staffers have primary degrees in journalism or media. The second-most represented subject area is English literature, with 4 degrees. The company’s only identifiable medical expert appears to serve NewsGuard in an advisory capacity and seems to hold a separate full-time job teaching science journalism in Italy. Other science and social science qualifications are sparse, and seldom surpass the undergraduate level. Most NewsGuard articles on Covid-19 topics and policies are written by Gregory, whose only identified qualification is a bachelor’s degree in Media Arts.
Noting the general deficit of scientific and social scientific expertise, AIER asked NewsGuard’s co-CEO Steven Brill if his company employed any individuals with advanced degrees in healthcare, medicine, economics, public policy, or related scientific and social scientific fields. Brill responded, “A lot of us have qualifications and graduate degrees in some of the fields you mention.” However the chart below belies his answer:
Figure III: NewsGuard Staff by Field and Highest Degree Attained:
For example, the aforementioned John Gregory would not qualify as an expert in most of the fields he is responsible for fact-checking. Before joining NewsGuard, his journalism background consisted primarily of writing news articles about the Orlando, Florida theme park industry. He also previously worked as a writer for TriMed Media, a for-profit consulting and marketing firm that publishes a series of newsletters targeted toward healthcare industry executives. Although TriMed’s focus area is topical to healthcare, it specifically promotes its newsletter products as trade and marketing publications for its clientele within the industry. Neither Gregory’s formal training nor his journalism experience appear to convey a reasonable claim to scientific expertise in medicine, healthcare policy, health economics, or related subject areas.
Of course, non-experts have every right to offer opinions on scientific and social-scientific matters. Whether or not they should be taken seriously as fact checkers or act as arbiters of scientific disputes is another question entirely.
Would NewsGuard Even Pass Its Own Tests?
If we’re going to be on this topic, we might as well check to see if NewsGuard is a reliable website by its own standards. Indeed, with its partnership with Microsoft and its roster of accomplished staff, the public should understand what kind of organization this is.
To test how NewsGuard holds up to its own rating system, we subjected its website and practices to the same criteria it uses to evaluate other sources. The results reveal a website that preaches a very different standard for others than it adheres to in its own work. The results are as follows:
“Does not repeatedly publish false content”
As documented, NewsGuard has an extensive track record of publishing false and misleading content, particularly related to Covid-19. The site admitted to making corrections to 21 separate articles where it had inappropriately labeled the lab leak hypothesis – since described as “plausible” and “warranting investigation” by mainstream scientists as well as President Biden – as a conspiracy theory. While these corrections are welcome, they are also ubiquitous and point to a pattern of NewsGuard writers jumping the gun to “fact check” claims that are not in error, but rather deviate from their own editorial positions. In the case of AIER and the GBD, NewsGuard had to issue corrections to two separate false claims within days of their article appearing. As of this writing, NewsGuard’s coverage of the GBD still includes uncorrected falsehoods about the GBD’s position on masks and misleading scientific assessments of its content by non-expert sources such as Matt Hancock. Although NewsGuard’s browser app subscription model precludes an independent audit of their fact-checking on other websites and topics, these problems suggest a wider pattern of errors that call into question the reliability of the company’s content.
“Gathers and presents information responsibly”
As documented, NewsGuard’s staff of fact checkers generally lack appropriate scientific and social-scientific qualifications to arbitrate factual and interpretive disputes in these subject areas. Instead, the company relies primarily on appeals to selectively curated external figures that they deem to be authorities on the relevant subject. When asked about how they select expert opinions, NewsGuard co-CEO Steven Brill told AIER that they always “quote real experts” and again that “we rely on – and quote — sources who are the experts.” Our review of NewsGuard’s analysis of the GBD, however, found that this was not the case. Instead of scientific experts, NewsGuard primarily relied on political figures such as Matt Hancock, including statements by Hancock that are well outside of the scientific consensus on the concept of herd immunity. When asked if they would be willing to correct Hancock’s erroneous statements, NewsGuard writer John Gregory defended his use of this source and did not respond to further follow-up questions.
An additional concern appears with how NewsGuard affixes its “warning label” system to published articles that its writers have not reviewed, let alone identified as failing a fact check. In AIER’s own experience with this company, we discovered that NewsGuard had affixed a “red” warning label to several articles we published about factual errors in the New York Times’s 1619 Project. NewsGuard and its writer John Gregory made no effort to evaluate the contents of these articles, and did not contact AIER for any comment on the application of this label to any subject other than its above-noted disputes over Covid-19 and the GBD. At the same time, NewsGuard applied a “green” label to the New York Times 1619 Project’s own website, despite never evaluating the content contained therein (the 1619 Project, it should be noted, contains significant uncorrected factual errors, and engaged in clear acts of journalistic misconduct to deflect criticism of its historical claims). This disparate treatment of the 1619 Project provides an example of misleading and inappropriate applications of rating labels to contents that are subject to legitimate factual disputes, but that NewsGuard itself did not examine or evaluate before attaching its label. As this example illustrates, NewsGuard’s practice of affixing rating labels to websites can create a false impression among its readers that the company has conducted a fact check on articles that it has not even evaluated.
Figure IV: A Comparison of NewsGuard’s ratings of the 1619 Project and AIER’s content on the 1619 Project, July 2021
“Regularly corrects or clarifies errors”
As noted, NewsGuard periodically corrects certain errors found in its published content. Recent corrections include the aforementioned edits to 21 articles on other website coverage of the lab leak hypothesis, and two corrections to factual misrepresentations about AIER and the GBD. At the same time however, NewsGuard appears to issue its corrections on an inconsistent basis. NewsGuard’s Gregory employed evasive wording to justify his continued deprecation of the lab leak hypothesis in other uncorrected articles, describing it as “not substantiated.” NewsGuard does not appear to apply similar descriptors to figures such as President Biden, Anthony Fauci, or mainstream journalism outlets such as the New York Times – all of which have recently lent credence to the plausibility of the lab leak hypothesis.
In AIER’s own case, NewsGuard declined to make an appropriate correction to misleading scientific claims about herd immunity by Matt Hancock. The company did not respond to further requests for corrections to its misrepresentation of the GBD, including the false claim that the Declaration opposes the use of face masks.
NewsGuard’s corrections policy does not appear to meet common standards for transparency. While NewsGuard claims to make timely and warranted corrections to its articles, these corrections are not independently verifiable without purchasing the company’s internet browser app and searching individual websites one at a time. Furthermore, the company does not appear to retain archived and accessible copies of its uncorrected articles for independent verification of the changes. NewsGuard did not respond to AIER’s inquiries about the lack of transparency arising from these unconventional practices.
“Handles the difference between news and opinion responsibly”
Despite imposing a strict standard for separating news from opinion on the other websites it rates, NewsGuard’s internal practices substantially blur these lines in its own content. The website claims that it has “no political axes to grind,” yet its affiliated personnel appear to be deeply involved in partisan and political causes. The company’s board of advisors includes multiple former elected officeholders and political appointees from the Bush and Obama administrations, as well as foreign governments. When asked about this apparent contradiction, NewsGuard co-CEO modified his descriptor of the company to “politically balanced.” In reference to one of his advisory board members, Brill further stated, “I’m curious what “ax” you think someone like Tom Ridge is grinding.” Ridge is the former Republican governor of Pennsylvania and Secretary of Homeland Security in the Bush administration (for reference, AIER made no specific claim about Ridge’s beliefs save to note that he is clearly a political figure, contrary to NewsGuard’s depiction of itself as a non-political company).
Brill himself also has longstanding political interests as an opinion commentator on healthcare policy, including the Affordable Care Act. His advocacy work in this area includes a 2015 book that proposes a “new vision of how we can fix American healthcare” through further addenda to the Obamacare program. When asked about the political connotations of his own writings on healthcare policy, Brill told AIER, “I’m a journalist. I write magazine articles and books that have offended people on both sides of an aisle.” This statement strongly suggests that Brill does not differentiate between editorial commentary and news reporting in his own work.
“Avoids deceptive headlines”
NewsGuard’s headline practices are responsible within the normal standards of journalism.
“Website discloses ownership and financing”
NewsGuard publishes a short list of “financial investors” on its website, but offers no indication of how much money they invested, what their budget is, what percentage of their financial support comes from outside investors, the nature of the contracts they maintain with companies such as Microsoft, or where their revenue comes from. When asked about their policy for handling conflicts of interest between investors and website ratings, NewsGuard co-CEO Steven Brill told AIER that “If there are, they are always disclosed prominently in our Nutrition Label” and offered an example of a disclosure of his co-executive Gordon Crovitz’s previous affiliation with the Wall Street Journal.
A review of NewsGuard’s ratings of several media websites reveals that Brill’s statement does not accurately reflect how his company handles several of its listed financial investors. NewsGuard lists former Reuters CEO Thomas Glocer among its financiers; however as of July 2021, Glocer’s investment in NewsGuard is not disclosed on the site’s rating of Reuters news service. NewsGuard similarly reports receiving investments from Nicholas Penniman IV, the retired publisher of the St. Louis Post-Dispatch. As of July 2021, Penniman’s investment is not disclosed on NewsGuard’s rating of the St. Louis Post-Dispatch. A list of insufficiently disclosed potential conflicts of interests and NewsGuard’s ratings may be found below:
Figure V: NewsGuard ratings of websites with undisclosed connections to its own financial investors, July 2021
NewsGuard Investor with past affiliation
Disclosure on NewsGuard Rating Website (July 2021)
Huffington Post Investigative Fund
Nicholas Penniman V – Former Executive Director & Cofounder
The American Prospect
Nicholas Penniman V – Former Associate Editor
Nicholas Penniman V – Former Publisher
Nicholas Penniman V – Former Editor
Center for Responsive Politics
Nicholas Penniman V – Current Board of Directors
St. Louis Post-Dispatch
Nicholas Penniman IV – Former Publisher (retired)
Arizona Daily Star
Nicholas Penniman IV – Former Senior VP
Thomas Glocer – Former CEO
Leo Hindery – Former Chairman
“Clearly labels advertising”
NewsGuard does not currently advertise on its website, except for its own for-purchase website browser application.
“Reveals who’s in charge, including possible conflicts of interest”
NewsGuard generally discloses its staff of writers, including the authors associated with individual articles. It does not however disclose how specific website ratings are assigned to its staff writers, or how NewsGuard decides whether a website will be subjected to its rating system.
As noted, NewsGuard’s health policy writer John Gregory was previously employed as a writer for TriMed Media, a for-profit industry consulting firm that publishes newsletters geared toward healthcare industry executives and offers a variety of website and marketing services “to boost awareness and revenue” for healthcare businesses. Gregory’s past affiliation with this company is disclosed on NewsGuard’s website; however, he describes himself as a “reporter” for this firm’s “news publications.” The potential for a conflict of interest seems apparent not only in NewsGuard’s coverage of healthcare topics, but also from a separate subscription-based “service” that NewsGuard offers to healthcare provider companies for the dissemination of its ratings.
When asked whether his previous connections to TriMed Media presented a potential conflict of interest for his coverage of healthcare issues, Gregory stated that TriMed Media’s consulting business was conducted by “the company’s separate marketing department, and I worked entirely on the editorial side of the operation.” He further stated, “They do not bias my ability to do my work at NewsGuard, as I did not perform any of those marketing department tasks when I worked at TriMed Media between 2016 and 2018.” At the time of this writing, TriMed Media’s website content does not appear to differentiate its news, editorial, or marketing/consulting departments or functions in the ways that Gregory describes in his response. TriMed Media’s website for its newsletter publications depicts their primary audience as for-profit healthcare industry executives. Gregory did not respond to follow-up questions asking him for elaboration on his relationship with and duties at TriMed Media. While Gregory’s case does evince disclosure of a past employer with potential conflicts of interest in his website ratings area, that disclosure remains sparse on details and offers an incomplete picture of the nature of TriMed Media’s business model. Furthermore, NewsGuard’s own subscription platform for healthcare providers illustrates that the company’s own business model operates in an overlapping area of the medical industry. Details on this service or how its subscription contracts operate are sparsely accounted for on NewsGuard’s website
“The site provides the names of content creators, along with either contact or biographical information”
NewsGuard’s articles generally list the identity of writers who contributed to their contents, and accompanying biographies include contact information for each writer.
NewsGuard applies a 100-point scorecard to the websites it rates, based on the above-listed criteria. Although their exact scoring rubric is only summarized on their website, we applied their publicly disclosed scoring system to each of the criteria above in light of our investigation.
Figure VI: Our “Nutrition Label” for the NewsGuard Website
Points Awarded to NewsGuard
Does not repeatedly publish false content
Gathers and presents information responsibly
Regularly corrects or clarifies errors
Handles the difference between news and opinion responsibly
Avoids deceptive headlines
Website discloses ownership and financing
Clearly labels advertising
Reveals who’s in charge, including possible conflicts of interest
The site provides the names of content creators
In light of all these considerations, it seems that NewsGuard fails to pass a number of important journalistic standards. Furthermore, it is concerning that NewsGuard cannot separate itself from its disagreement with the content of AIER’s publications regarding public health, which dominates its report. This is even more concerning when considering its dubious track record to responsibly or reliably use its self-appointed position of power to police the very information it purports to be the arbiter of truth on. Furthermore, it is the supreme height of arrogance to claim a monopoly on scientific truth, not just without a background in science or really any field outside of journalism, but during a novel issue with constantly evolving information.
Our Rating of NewsGuard: 36.25/100.
This website fails to adhere to several basic journalistic standards, and should be used with extreme caution as a source for verifying the reliability of the websites it purports to rate.
The truth is best sought through the marketplace of ideas where reason and evidence are the weapons of choice. When we see fact checkers like NewsGuard, who not only fail to uphold their high-sounding principles but even publicly encourage working with the government to suppress speech, we should raise red flags. NewsGuard is clearly illustrative of the now tired idea that Covid-19 showed us that we should simply just do as we’re told and don’t question the government or its experts. On this matter, they have shown themselves to be either unable to appropriately moderate public discourse or act as little more than cheerleaders for favored political figures and their preferred policy approaches to Covid-19. It wouldn’t be a stretch if they happen to be both.