Why is Meta labelling Trump’s Real mug shot as a fake?
Click here to listen to an audio version of this post, read by the author. I recently blogged about how Instagram was blocking a large list of hashtags specifically related to the American Democratic Party. Dozens of tags like #Democrat, #Biden, #Kamala, and #
... show more
Why is Meta labelling Trump’s Real mug shot as a fake?
Click here to listen to an audio version of this post, read by the author. I recently blogged about how Instagram was blocking a large list of hashtags specifically related to the American Democratic Party. Dozens of tags like #Democrat, #Biden, #Kamala, and #DNC were being censored, but Republican tags were not affected. Today I found another weird issue that I thought was worth documenting here. Whenever users post Donald Trump’s booking photo from the Fulton County Jail, Meta adds a “fact check” beneath the post claiming the photo is “not real” and “made using AI”. I saw other users pointing this out, so I gave it a try too.
When I posted Trump’s mug shot on Threads, I got a notification saying a “fact check” had been added to my post. The note appeared under my post and said “see context added to this post”. When you clicked on the post you’d see a message saying, “Fact Check: Trump ‘Mug Shot’ Is NOT Real — It Was Made Using AI” along with a link to a site called Lead Stories. The story was written April of 2023, a few months before the real mug shot of Trump was taken. It is unrelated to my post and includes a totally different image than the one I shared.
I filed a “dispute” on the fact check and received a notification 20 minutes later that my dispute had been denied and the label would remain under my post. My post was later shared by a few influential users on Threads, including someone who had recently been tagged in a post from Zuckerberg himself. And on the next morning, I found that the added context had been removed from my post with no explanation or notifications from Meta. It seems that my post’s visibility drew attention from someone at Meta, who removed the fact check from my specific post, but not from any others. As of this writing, the note still appears under a post by @TeddyHose.
The problem Persists:
To confirm this is still happening, I just shared Trump’s mug shot on Facebook and immediately got a notification that the image is still being incorrectly labelled by Meta’s moderation system. The Facebook notification specifies that my “post includes a photo that independent fact-checkers say was altered.” Other folks have confirmed this too, and I encourage you to try sharing Trump’s mug shot photo to see if it’s still happening too.
My thoughts:
Meta’s fact check on my post takes you to a link from early 2023 that includes a photo that looks totally different from the image that I shared. The article was accurate at the time it was written, but statements like “No credible outlet has published a police mugshot of Trump.” were no longer accurate a few months later.
It appears that Meta has a big database of images that always get flagged with a fact check. For some reason, an employee added Trump’s real mug shot to this database and connected that real photo with a real, but unrelated news article. Thus blurring the lines of what is and isn’t real on these platforms. It’s hard to understand how such a glaring error could take place and why it hasn’t been corrected in the past 18 months, but the end result is the same.
Again, the two photos look nothing alike, and Meta has had years to correct this. For added context, keep in mind this issue is happening
18 days after Zuckerberg said he was ending Meta’s fact-checking program
During the first week of Trump’s second term
Shortly after Meta was caught censoring all of the most popular hashtags related to Democrats on Instagram
Last year, Facebook also added incorrect labels on articles that were critical of Project 2025. Even if the article was accurate, Facebook added a message saying the articles were false and cited the Dispatch, a conservative fact-checking group, as their source.
Mark Zuckerberg claims that Meta shouldn’t be in the business of fact-checking and that community notes would be a better alternative. I’d say he’s right in this particular instance, but if you take these erroneous fact checks along with other recent incidents, like censoring all hashtags related to Democrats, I’d say Zuckerberg is really making a case for why we should be ditching Facebook, and using decentralized social media alternatives like Mastodon and Pixelfed.
Listen to an audio version of this long-ass post. On the day of Trump’s second presidential inauguration, Instagram censored search results for the most popular hashtags associated with American Democrats, like “#Democrat”, “#PresidentBiden”, “#VoteBlue”, “#MichelleObama”, “#Kamala”, and dozens more. This ban lasted for 9 total hours and affected users worldwide, regardless of their content settings. During that same time, Instagram was displaying full search results for “#Republican”, “#Trump”, “#VoteRed”, “#PresidentTrump”, and every other popular Republican hashtag I tried. I found about 50 censored political tags related to Democrats, but every Republican tag was still working fine. The removal of these pro-Democrat hashtags started on the evening of January 20th, 2025, coincidentally a few hours after Trump signed the “Restoring Freedom Of Speech” executive act. These censored results lasted for about 9 hours, then went back to normal at around 5 am PST. I want to state again that these words were censored for all Instagram users on the website and app regardless of that user’s content settings or their location. Meta claimed this was a technical “error” affecting hashtags “across the political spectrum,” but this is demonstrably false. I searched through the most popular hashtags used by Democrats and Republicans last night and found that the majority of Democrat terms were being censored, while none of the popular Republican tags were affected. This selective censorship is a scary look into Meta’s updated content moderation practices and its potential alignment with the new administration.
The following hashtags were actively blocked by Instagram on the first day of Trump’s second presidency. While they are now searchable again, I personally confirmed all of these hashtags were inaccessible last night:
Barack
BarackObama
Bernie
BernieSanders
Biden
Constitution
Democrat
Democrats
DNC
Fascism
FuckTrump
Insurrection
Jan6th
JohnOliver
JoeBiden
Kamala
Leftist
Liberal
Liberals
MichelleObama
Obama
OccupyDemocrats
PresidentBiden
PresidentObama
Queer
ReproductiveRights
VoteBlue
And those were just the ones I could find yesterday. I’m sure there are many more. Instagram is still blocking other hashtags like ProChoice, GeorgeConway (a writer critical of Trump), JackSmith (the special counsel against Trump), 14thAmendment, and Drumpf. It’s possible those words were already being censored before Trump’s inauguration, so I didn’t include them on the list above.
During that same 9-hour window, I confirmed the following words were not being censored on IG:
AltRight
DonaldJTrump
DonaldTrump
FuckObama
FuckBiden
FuckKamala
LetsGoBrandon
MelaniaTrump
PresidentTrump
Republican
RNC
Trump
VoteRed
Those words all worked. You couldn’t search for FuckTrump, but you could search for FuckObama or FuckBiden. Trump was searchable, but Biden, Kamala, Bernie, and Obama were not. You could search for MelaniaTrump, but not MichelleObama. PresidentTrump still showed millions of results, while discussions of PresidentBiden were marked as “sensitive content”.
We don’t know what caused Meta to censor these words, but we do know this was not affecting “both sides.”. This was an extremely targeted list of censored words.
As a contrast, I saw several claims on Threads and IG that TikTok was also heavily censoring content. I tested it out for myself and was still fully able to search for words like “Fascist,” “Fascism,” “Democrat,” and “Trump”. I was also able to post pro-Palestinian content like “Free Palestine” without having the post removed. If TikTok is censoring content, they are being more clever about it than Instagram is.
Though we can prove that Instagram was censoring certain hashtags last night, we can’t say why they were doing it. It’s possible this really was a computer error with no human oversight, but I’ve never heard of a computer bug that only affects people based on their political affiliation.
To understand why this happened, take a look at the context surrounding it. Last year, Trump accused Zuckerberg of election interference and threatened to lock Mark up “for life”. After Trump won the election, Zuckerberg started a series of regular visits to Mar-a-Lago for private meetings with the President-Elect. Last week, Zuckerberg announced big changes to Meta’s content moderation team and added Dana White (a Trump loyalist) to Meta’s board. Yesterday, Zuckerberg attended Trump’s inauguration along with Jeff Bezos, Elon Musk, Tim Cook, Sundar Pichai, Shou Zi Chew, and several other of the richest tech oligarchs.
With all of that in mind, I find it unlikely that an Instagram error selectively silencing Trump’s political opponents would coincidentally occur on the first day of his second presidency. Given the state of *gestures broadly at everything around me*, I believe that Meta has many systems in place to quickly silence Trump’s political opponents, and last night we saw one of those systems being used. If Meta really did have an “error” yesterday, it was probably that they activated that system too early. #Biden#Obama#Kamala#Trump#BarackObama#DonaldTrump#MichelleObama#MelanieTrump#Democrat#Republican#leftist#AltRight#Liberal#conservative#OccupyDemocrats#Maga#VoteBlue#GOP#PresidentBiden
Listen to an audio version of this long-ass post. On the day of Trump’s second presidential inauguration, Instagram censored search results for the most popular hashtags associated with American Democrats, like “#Democrat”, “#PresidentBiden”, “#VoteBlue”, “#
... show more
The Day Instagram Blocked Democracy
Listen to an audio version of this long-ass post. On the day of Trump’s second presidential inauguration, Instagram censored search results for the most popular hashtags associated with American Democrats, like “#Democrat”, “#PresidentBiden”, “#VoteBlue”, “#MichelleObama”, “#Kamala”, and dozens more. This ban lasted for 9 total hours and affected users worldwide, regardless of their content settings. During that same time, Instagram was displaying full search results for “#Republican”, “#Trump”, “#VoteRed”, “#PresidentTrump”, and every other popular Republican hashtag I tried. I found about 50 censored political tags related to Democrats, but every Republican tag was still working fine. The removal of these pro-Democrat hashtags started on the evening of January 20th, 2025, coincidentally a few hours after Trump signed the “Restoring Freedom Of Speech” executive act. These censored results lasted for about 9 hours, then went back to normal at around 5 am PST. I want to state again that these words were censored for all Instagram users on the website and app regardless of that user’s content settings or their location. Meta claimed this was a technical “error” affecting hashtags “across the political spectrum,” but this is demonstrably false. I searched through the most popular hashtags used by Democrats and Republicans last night and found that the majority of Democrat terms were being censored, while none of the popular Republican tags were affected. This selective censorship is a scary look into Meta’s updated content moderation practices and its potential alignment with the new administration.
The following hashtags were actively blocked by Instagram on the first day of Trump’s second presidency. While they are now searchable again, I personally confirmed all of these hashtags were inaccessible last night:
Barack
BarackObama
Bernie
BernieSanders
Biden
Constitution
Democrat
Democrats
DNC
Fascism
FuckTrump
Insurrection
Jan6th
JohnOliver
JoeBiden
Kamala
Leftist
Liberal
Liberals
MichelleObama
Obama
OccupyDemocrats
PresidentBiden
PresidentObama
Queer
ReproductiveRights
VoteBlue
And those were just the ones I could find yesterday. I’m sure there are many more. Instagram is still blocking other hashtags like ProChoice, GeorgeConway (a writer critical of Trump), JackSmith (the special counsel against Trump), 14thAmendment, and Drumpf. It’s possible those words were already being censored before Trump’s inauguration, so I didn’t include them on the list above.
During that same 9-hour window, I confirmed the following words were not being censored on IG:
AltRight
DonaldJTrump
DonaldTrump
FuckObama
FuckBiden
FuckKamala
LetsGoBrandon
MelaniaTrump
PresidentTrump
Republican
RNC
Trump
VoteRed
Those words all worked. You couldn’t search for FuckTrump, but you could search for FuckObama or FuckBiden. Trump was searchable, but Biden, Kamala, Bernie, and Obama were not. You could search for MelaniaTrump, but not MichelleObama. PresidentTrump still showed millions of results, while discussions of PresidentBiden were marked as “sensitive content”.
We don’t know what caused Meta to censor these words, but we do know this was not affecting “both sides.”. This was an extremely targeted list of censored words.
As a contrast, I saw several claims on Threads and IG that TikTok was also heavily censoring content. I tested it out for myself and was still fully able to search for words like “Fascist,” “Fascism,” “Democrat,” and “Trump”. I was also able to post pro-Palestinian content like “Free Palestine” without having the post removed. If TikTok is censoring content, they are being more clever about it than Instagram is.
Though we can prove that Instagram was censoring certain hashtags last night, we can’t say why they were doing it. It’s possible this really was a computer error with no human oversight, but I’ve never heard of a computer bug that only affects people based on their political affiliation.
To understand why this happened, take a look at the context surrounding it. Last year, Trump accused Zuckerberg of election interference and threatened to lock Mark up “for life”. After Trump won the election, Zuckerberg started a series of regular visits to Mar-a-Lago for private meetings with the President-Elect. Last week, Zuckerberg announced big changes to Meta’s content moderation team and added Dana White (a Trump loyalist) to Meta’s board. Yesterday, Zuckerberg attended Trump’s inauguration along with Jeff Bezos, Elon Musk, Tim Cook, Sundar Pichai, Shou Zi Chew, and several other of the richest tech oligarchs.
With all of that in mind, I find it unlikely that an Instagram error selectively silencing Trump’s political opponents would coincidentally occur on the first day of his second presidency. Given the state of *gestures broadly at everything around me*, I believe that Meta has many systems in place to quickly silence Trump’s political opponents, and last night we saw one of those systems being used. If Meta really did have an “error” yesterday, it was probably that they activated that system too early. #Biden#Obama#Kamala#Trump#BarackObama#DonaldTrump#MichelleObama#MelanieTrump#Democrat#Republican#leftist#AltRight#Liberal#conservative#OccupyDemocrats#Maga#VoteBlue#GOP#PresidentBiden
If you think #Trump really did #Jan6, but you refuse to take the very real, very believable issues around the #Biden presidency seriously, you're just an empty gong.
I didn't vote Trump either time, but anyone can see he's being railroaded.
Come on infosec people, you *know* the signs of social engineering/phishing. High stakes, time constraints, call to action. You're not being smart, you're being a mark.
If you cannot articulate your opponent's position as they would articulate it, you are not qualified to have an opinion on that person or able to have a rational debate. Reductive malicious catch phrases like "Don't Say Gay BIll" or "Republicans want to ban boo
... show more
If you think #Trump really did #Jan6, but you refuse to take the very real, very believable issues around the #Biden presidency seriously, you're just an empty gong.
I didn't vote Trump either time, but anyone can see he's being railroaded.
Come on infosec people, you *know* the signs of social engineering/phishing. High stakes, time constraints, call to action. You're not being smart, you're being a mark.
If you cannot articulate your opponent's position as they would articulate it, you are not qualified to have an opinion on that person or able to have a rational debate. Reductive malicious catch phrases like "Don't Say Gay BIll" or "Republicans want to ban books" are great slogans for ginning up emotional reactions, but don't describe the intent or heart behind the actions. Exactly -zero- Republicans I've ever met want to actually ban books. I've run into more on the Left who want to control speech than vice versa.
I'm horrified by the idea of either the swamp things on the Republican side or the Comrades in the Democrat party taking complete control. The result is the same - "A boot standing on a human face - forever."
If you have student loans, check out the new SAVE plan from the Biden administration. It cuts payments to 5% of disposable income (income in excess of 225% of the poverty line) with many households qualifying for $0 payments.
Getting student loans down to 5% of income is a huge deal!
"On #right-wing forums, supporters of President #Trump are openly planning attacks on state capitols for the days leading up to President-elect Joseph R. #Biden Jr.'s inauguration on Jan. 20."
