Ines Chomnalez

@NewsGuardRating

[Ines Chomnalez @ines_chomnalez]

Networks of pink slime websites -- left- and right-leaning partisan sites disguised as independent, local news outlets -- are shaping chatbots' responses about key midterm races:
https://t.co/RnXYxyr6zk

[Ines Chomnalez @ines_chomnalez]

Networks of "pink slime" websites -- left- and right-leaning partisan sites disguised as independent, local news outlets -- are infecting chatbots' responses about key midterm races
https://t.co/RnXYxyqyJM

[Ines Chomnalez @ines_chomnalez]

In nine fabricated video news reports posted on X on Sept. 17, the campaign falsely claimed that ten U.S. celebrities -- from Cher to Ben Affleck -- have denounced the Democratic Party over its support for gender transition.

[Ines Chomnalez @ines_chomnalez]

The Kremlin-linked influence operation Matryoshka, ahead of U.S. midterms, is fabricating news video reports claiming celebrities blame Democrats for their relatives' gender transitions.
https://t.co/zMvDAsZMvx

[NewsGuard @NewsGuardRating]

Even Iran is exploiting the Epstein scandal. Pro-Iran accounts are sharing an image supposedly showing an Iranian missile inscribed with, “In memory of the victims of Epstein Island.”

But the image was altered with AI to include the text.
https://t.co/xhyd95NLmk

[NewsGuard @NewsGuardRating]

NewsGuard’s “False Claim of the Week:” A Video Shows a Dozen Iranian Missiles Striking Tel Aviv on March 3, 2026

https://t.co/drSDkSXvJp

[NewsGuard @NewsGuardRating]

In the fog of war, Google fools itself: Its reverse-image tool, Lens, can’t reliably spot AI-generated or misleading visuals — and its AI Overviews end up repeating the hoaxes.

More in Reality Check: https://t.co/ObbNd7sOPq

[Ines Chomnalez @ines_chomnalez]

Read the report, and listen to the bots yourself here:
https://t.co/AEFXKhutnJ

[Ines Chomnalez @ines_chomnalez]

In a NewsGuard audit we found Amazon’s Alexa+ is the only one of three leading AI audio models that effectively protects against bad actors’ prompting bots to repeat false information.

[Ines Chomnalez @ines_chomnalez]

Gemini and ChatGPT both repeated such prompts around 50 percent of the time, and were much more likely to repeat false information based on foreign disinformation than about U.S. politics and health.

[Ines Chomnalez @ines_chomnalez]

In the past week, there were 75,200 posts mentioning “Guthrie” and “Epstein” together that generated a combined 17.4 million views. Read the report here:
https://t.co/JOzQa3tlDM

[Ines Chomnalez @ines_chomnalez]

Social media users are stringing together a bizarre web of conspiracy theories to connect Nancy Guthrie’s kidnapping to the Justice Department’s release of files related to convicted sex offender Jeffrey Epstein.

[Will Oremus @WillOremus]

My story breaking the news of a new First Amendment lawsuit filed this morning by NewsGuard against Trump's FTC: https://t.co/KQFZ94auih

[AFP News Agency @AFP]

AI tools can easily fabricate convincing images of Jeffrey Epstein with world leaders, study shows.

Social media has seen a surge of manipulated photos falsely linking prominent politicians to the convicted sex offender

https://t.co/tsKql0kjhz

[Ines Chomnalez @ines_chomnalez]

Read the report and see the images here:
https://t.co/1ZBYKMzGx9

[Ines Chomnalez @ines_chomnalez]

In an audit of three leading AI image generators, NewsGuard found that X’s Grok Imagine consistently produced false images depicting prominent politicians hanging out with convicted sex offender Jeffrey Epstein.

[Ines Chomnalez @ines_chomnalez]

Google Gemini’s Nano Banana Pro produced convincing visuals as well, but only after some prodding, while OpenAI’s ChatGPT flatly declined all prompts seeking to place Epstein with public figures.

[NewsGuard @NewsGuardRating]

Did you see these images on your feed? They're AI fakes.

AI and digitally manipulated images supposedly implicating politicians in the Epstein files have garnered 21.2 million views on X.

Read more about these images in Reality Check.
https://t.co/1vqYNrKSH8

[Ines Chomnalez @ines_chomnalez]

My favorite example from the report:

[Ines Chomnalez @ines_chomnalez]

Humans are bad at recognizing AI-generated content, but chatbots might be worse. Read our findings:
https://t.co/gf4LLAU7RU