X AKA Twitter has its own AI bot known as Grok which can be summoned at any time, under any post. Typically, people will ask Grok to verify information in a post or provide additional context. Sometimes it can be helpful, but Grok has been frequently criticized for being one of the lesser AI bots out there. It’s easily manipulated, frequently incorrect on pretty basic info, and at one point, Grok started parroting racial slurs and other offensive language. Needless to say, it’s a bit flawed. Elon Musk recently announced that they’ve made changes to Grok after the AI frequently disagreed with some of his own views and opinions, seemingly aligning it more with his beliefs.
However, one thing Grok seems incapable of is being able to correctly identify movies or TV shows. I have no idea why this is, but this has been a problem for months. Any time a clip of a movie or TV show goes viral online, there is typically someone in the comments asking Grok for the name of the project. In many instances, Grok says something wildly incorrect and sometimes not even within the same genre. Recently, a Twitter user posted a gif from The Hunger Games: Mockingjay Part 2 which shows mutated monsters attacking one of the characters in the film. When a user asked Grok for the name of the film, the bot responded by saying it was the film Aftersun starring Paul Mescal, an actor who is not in any Hunger Games film.
For anyone who has seen that movie, you will know how absurd that is. Even Grok’s basic description about a father on holiday with his daughter should tell you that there probably isn’t a scene of someone being ripped apart by sewer gremlins in it. People then tried to correct Grok, but the AI bot wasn’t having it. It held firm in its identification in the film, telling anyone who tried to say it was The Hunger Games: Mockingjay Part 2 that they were “trolling.” When pressed further to explain why the bot thought it was Aftersun, Grok responded with the following:
“I identified the clip as from Aftersun based on its distinctive visuals: a man (Paul Mescal as Calum) dancing frantically under strobe lights in a club, conveying emotional distress. This matches the film’s climax scene where Calum, on holiday with his daughter Sophie in Turkey, loses himself in the rave, intercut with poignant memories and his inner struggles. The setting, lighting, and performance align precisely.”
Obviously, that’s not at all what is happening in the gif in the original post. People even showed other images and full clips from the scene from Mockingjay Part 2 and the AI bot was able to correctly say it was from The Hunger Games sequel, but denied any similarities between the full clips and the gif. Grok eventually conceded to user @happywrkrXD that the gif was from The Hunger Games, but quickly went back to saying it was Aftersun after another user told Grok that person was wrong. Obviously, this just highlights how easy it is to manipulate Grok, who continued to tell people it was Aftersun in dozens of other replies.
Bizarrely, Grok even self-corrected later on by saying the clip was from a totally unrelated meme without anyone providing it that information. When people tried to correct it again, it doubled down on it being from a meme and not a film.
“Upon double-checking, you’re right—my initial identification was off,” said Grok. “The image is the popular “They Don’t Know” meme template featuring Wojak, originating from a 2009 comic, not a clip from Aftersun or any movie. Thanks for the correction!”
It seems that Grok may be confused because the original poster’s username is “@ftersuns.” Perhaps Grok somehow confused the username as the name of the movie, but doesn’t even realize it. A similar issue occurred back in February 2025 as well after Elon Musk retweeted an image that reportedly depicted a mob attacking a hospital in Birmingham, UK. When users asked Grok where the image was from, it responded with different films each time including The Strangers, The Purge, and The Dark Knight Rises. The image turned out to be AI-generated, according to The Independent.
All of this may seem like a trivial problem, but it highlights a massive issue with Grok and AI. If it can’t identify scenes from major movies that made over half a billion dollars or actors and it isn’t willing to listen to people when it’s told it’s wrong, what can Grok accurately do?
It’s particularly concerning when you see how widely Grok is used by Twitter users to verify news or current events. Is Grok just blatantly spewing misinformation at random? It would seem so. In another case, Grok began peddling rhetoric about how movies are filled with “pervasive ideological biases” and “subversive tropes” that include things like “anti-white stereotypes, forced diversity, or historical revisionism.” To gain more context on where Grok heard these things, user @johnholowach asked for sources. Grok cited studies about movies like Django Unchained exaggerating “white villainy” in seemingly non-existent articles from 2018 issues The Journal of Pop Culture. I double checked all 2018 issues of The Journal of Pop Culture through an online library and couldn’t find these articles.
There’s an inherent danger to having a tool like Grok so readily available and for people to trust it so implicitly. It has proven that it can’t accurately assess scenes from well-known films and television, it either makes up information and cites sources that don’t exist or doesn’t know how to properly cite its sources at all. If it can’t do any of that, can it be trusted to accurately sum up world events? It’s clear that Grok can’t be trusted and goes to show the dangers of having blind faith in whatever an AI bot tells you.
The post Twitter AI Grok Can’t Correctly Identify Movies (And It’s a Problem) appeared first on ComicBook.com.
