Skip to main content
Ray Bogan Political Correspondent
Share
Politics

House GOP probes Meta, Google over Trump assassination attempt ‘censorship’

Share
Ray Bogan Political Correspondent
Share

In the wake of the assassination attempt against former President Donald Trump, some strange and inaccurate things happened on the internet. Meta’s AI chatbot said the assassination attempt was a fictional event, Facebook fact-checked and censored the iconic image of Trump raising his fist, and Google’s autofill search feature produced results for the failed assassination attempts on Harry Truman and Ronald Reagan, or the historic assassination of Archduke Franz Ferdinand, but not Trump. 

Media Landscape

MediaMiss™This story is a Media Miss by the left as only 0% of the coverage is from left leaning media. Learn more about this data
Left 0% Center 20% Right 80%
Bias Distribution Powered by Ground News

The companies admitted to the shortcomings and provided clear explanations. However, the House Oversight Committee wants to know more, and it’s opening an investigation into how these products are designed.

“On behalf of the American people, the Committee is dedicated to fully understanding when and how information is being suppressed or modified, whether it be due to technical error, a policy intended to ensure safety, or a specific intent to mislead,” Chairman James Comer, R-Ky., said in a statement. 

QR code for SAN app download

Download the SAN app today to stay up-to-date with Unbiased. Straight Facts™.

Point phone camera here

Comer is asking Meta and Google for all internal policies and documents related to how their AI chatbot and search products are designed, trained, reviewed, managed and updated.

The committee want to know how these products take digest recent newsworthy events and how they preclude or limit information the companies view as harmful or dangerous. 

Here’s what prompted the investigation. 

Immediately after the assassination attempt, Meta’s AI chatbot told users it was a fictitious event. Meta’s vice president of U.S. Global Policy, Joel Kaplan, explained in a blog post what happened. 

“Rather than have Meta AI give incorrect information about the attempted assassination, we programmed it to simply not answer questions about it after it happened and instead give a generic response about how it couldn’t provide any information,” Kaplan wrote. 

“In a small number of cases, Meta AI continued to provide incorrect answers, including sometimes asserting that the event didn’t happen,” he added. “These types of responses are referred to as hallucinations, which is an industry-wide issue we see across all generative AI systems, and is an ongoing challenge for how AI handles real-time events going forward.” 

Facebook also censored and fact-checked the picture of Trump holding up his fist because a user posted a doctored image of Secret Service agents smiling as they surrounded him. A fact-check label was added to the fake image which caused the technology to scan the site for all posts that contained it. Because the real image is nearly identical, the system added the fact-check to it. 

Finally Google’s autofill didn’t complete searches for the assassination attempt, instead it provided information on other presidents or Archduke Franz Ferdinand. 

According to the committee staff inquiries, “Google contends that the autocomplete results omitted the Trump assassination attempt due to a safety protocol concerning predicted assassination attempts of current political leaders, and Google had not yet updated the autocomplete feature to reflect that an assassination attempt of President Trump had occurred.”

Tags: , , , , , , , , , ,

[RAY BOGAN]

In the wake of the assassination attempt against former President Trump, some strange and inaccurate things happened on the internet. 

Metas AI chatbot started telling people the assassination attempt was a fictional event, Facebook fact checked and censored the iconic image of Trump raising his fist, Google’s autofill search feature produced results for the failed assassination attempts on Harry Truman and Ronald Reagan, or the historic assassination of Archduke Franz Ferdinand, but not Trump. 

The companies have admitted to the shortcomings and provided clear explanations for why all that happened. But the House Oversight Committee wants to know more and is opening an investigation into how these products are designed. 

Republican Chairman James Comer said in a statement – “On behalf of the American people, the Committee is dedicated to fully understanding when and how information is being suppressed or modified, whether it be due to technical error, a policy intended to ensure safety, or a specific intent to mislead. 

Comer is asking Meta and Google for all internal policies and documents related to how their AI chatbot and search products are designed, trained, reviewed, managed and updated. They want to know how these products take into account recent newsworthy events and how they preclude or limit information the companies view as harmful or dangerous. 

Here’s what prompted the investigation. 

Immediately after the assassination attempt, Metas AI chatbot told people it was a fictitious event. 

In a blog post, Meta’s VP of Global Policy Joel Kaplan explained: “Rather than have Meta AI give incorrect information about the attempted assassination, we programmed it to simply not answer questions about it after it happened – and instead give a generic response about how it couldn’t provide any information.”

He went on to say, “In a small number of cases, Meta AI continued to provide incorrect answers, including sometimes asserting that the event didn’t happen…” 

“These types of responses are referred to as hallucinations, which is an industry-wide issue we see across all generative AI systems, and is an ongoing challenge for how AI handles real-time events going forward.” 

Facebook also censored and fact checked this picture of Donald Trump holding up his fist. That’s because someone posted a doctored image of the Secret Service agents smiling as they surrounded him. A fact check label was added to the fake image which caused their technology to automatically scan the site for all posts that contained it. Because the real image is nearly identical, their system added the fact check to it. 

Finally Google’s autofill didn’t complete the Donald Trump assassination attempt, instead it provided information on other presidents or Archduke Franz Ferdinand. 

According to committee staff inquiries – “Google contends that the Autocomplete results omitted the Trump assassination attempt due to a safety protocol concerning predicted assassination attempts of current political leaders, and Google had not yet updated the Autocomplete feature to reflect that an assassination attempt of President Trump had occurred.” 

STraight ARrow News will follow this investigation, so for more reporting straight from capitol hill download the straight arrow news app.