A months-long Media Research Center investigation found that Apple News and Google News failed to surface a single negatively framed headline about Michigan Democratic Senate candidate Abdul El-Sayed after February, raising fresh questions about algorithmic bias just days before his primary.
MRC researchers tracked the top 20 stories displayed each morning on both platforms from early November through July 21, logging results at 8:30 a.m. ET on a computer wiped of search histories and cookies before every session. Their conclusion: at least 104 stories from right-leaning outlets covering El-Sayed's record, policy positions, and public statements never appeared in the feeds they monitored. After February 10, not one headline in those top-20 results framed El-Sayed negatively, Fox News Digital reported.
The findings land less than two weeks before Michigan's August 4 Democratic Senate primary, where El-Sayed faces Rep. Haley Stevens in a two-candidate race. State Sen. Mallory McMorrow recently suspended her campaign, and Gov. Gretchen Whitmer endorsed Stevens on Friday, moves that sharpened the stakes and made the information environment around El-Sayed all the more consequential for voters.
MRC President David Bozell framed the study as part of a broader pattern. The organization previously investigated how the same platforms handled coverage of Graham Platner, a former Maine Senate candidate who was forced to exit his race after a rape accusation from a former girlfriend. That earlier probe found at least 112 stories scrutinizing Platner went unpromoted by Apple News and Google News between November and May.
Bozell told Fox News Digital that Google News failed to feature more than 250 stories covering controversies involving both El-Sayed and Platner combined. The outlet imbalance was stark: 69 percent of the outlets appearing in Google News' top selections were left-leaning, the MRC found, while just 3 percent were right-leaning.
That ratio matters because of how most people actually consume news. Bozell put it plainly:
"Most Americans don't search dozens of news websites every day. They open Apple News or Google News and read what's presented to them."
If the platforms' algorithms, or editorial choices, tilt the playing field, the average voter never sees the stories that might change a mind. The voter does not know what is missing.
The stories that went missing were not trivial. El-Sayed supports Medicare for All, abolishing Immigration and Customs Enforcement, and shifting funding away from police departments. In April, CNN's Manu Raju asked him whether he considered the Israeli government to be as evil as Hamas. El-Sayed did not hedge:
"Yes, killing tens of thousands of people makes you pretty d*** evil. Hamas: Evil, Israeli government: Evil. We can say both."
He also campaigned with Hasan Piker, a left-wing streamer known for inflammatory remarks about the September 11 attacks, and accepted an endorsement from Osama Siblani, leader of the Arab American Political Action Committee. Questions were raised about El-Sayed's use of the physician title as well. Fox News Digital reached out to El-Sayed for comment and did not immediately receive a response.
These are the kinds of stories that, in a functioning news ecosystem, would circulate widely enough for voters to weigh before casting a ballot. The MRC's data suggests they circulated almost nowhere on the two largest news aggregation platforms in the country.
The pattern fits a broader institutional tendency to shield preferred narratives from scrutiny, whether through editorial gatekeeping, algorithmic design, or simple neglect of inconvenient reporting.
Anticipating pushback, the MRC designed its tracker to address the most obvious critique: that algorithmic feeds are personalized, so one user's results prove nothing. Researchers used a clean computer, cleared cookies and search history before every session, and examined Google News' default home tab rather than its search engine. They searched for "Michigan," "Abdul," and "El-Sayed." Before launching the tracker, the team tested whether sources and story types changed throughout the day and found no statistically significant differences, which is why they settled on the single 8:30 a.m. snapshot.
Apple News and Google News each featured only one headline about the Michigan Senate race during the entire tracking period. Neither headline focused on controversies involving El-Sayed.
Google has pushed back on the MRC's approach before. A Google spokesperson called the claims from the earlier Platner study "totally false and based on a completely flawed methodology," arguing that it relied on a single daily snapshot from one account and failed to account for updates, interests, and location. Fox News Digital reached out to both Apple and Google for comment on the El-Sayed findings and did not immediately receive responses.
Apple has previously said its Top Stories are selected by editors and appear the same for every user, while other portions of its Today feed can be personalized. That distinction only deepens the question: if human editors pick the Top Stories and still chose zero negatively framed El-Sayed headlines after February 10, the problem is not an algorithm. It is an editorial decision.
The MRC study arrives in a political climate where trust in institutions that handle information, tech platforms, intelligence agencies, legacy media, is already fractured. Americans watched a coordinated effort to discredit the Hunter Biden laptop story before the 2020 election, an operation that relied on the same gatekeeping architecture: platforms suppressed the story, legacy outlets dismissed it, and voters went to the polls with incomplete information.
Bozell drew the connection explicitly. He said Google News "presents itself as a comprehensive news product" but leaves consumers wondering "whether major, newsworthy reporting about candidates is being left out of what they see."
"Our findings raise serious questions about whether Google News is providing users with the complete picture they expect."
The MRC's numbers do not prove intent. They do not explain why the suppression happened. But the scale, 104 missing stories on one candidate, 250-plus across two races, a 69-to-3 percent outlet tilt favoring left-leaning sources, describes a system that produces a consistent, one-directional result.
Whether that result flows from deliberate editorial choices, baked-in algorithmic assumptions, or some combination, the effect on voters is the same. They see what the platforms show them. And what the platforms showed them, for more than five months, was a Michigan Senate candidate with no visible controversies.
Accountability for that kind of information asymmetry has been slow to materialize. Institutions that shape public knowledge, from federal agencies dismissing internal dissenters to tech platforms curating what millions of voters read each morning, face little consequence when their outputs consistently favor one political direction.
El-Sayed, meanwhile, heads into the August 4 primary against Stevens with the benefit of an information environment that, according to the MRC's data, has largely insulated him from critical coverage on the platforms where most casual news consumers spend their time. Whether that insulation holds through the final days of the race may depend on whether voters encounter reporting outside the algorithmic bubble, and whether the platforms bother to explain the gap.
The pattern has become familiar enough to name: institutions that control the flow of politically relevant information keep producing outcomes that benefit one side. The question is no longer whether it happens. It is whether anyone with the power to fix it will be held accountable for letting it continue.
Voters deserve a complete picture. When the platforms that deliver their news decide, by algorithm or by hand, which stories to bury, they are not curating information. They are choosing sides.