Keep up with the innovative tech transforming business
Tech Brew keeps business leaders up-to-date on the latest innovations, automation advances, policy shifts, and more, so they can make informed decisions about tech.
As if the US presidential election needs more twists, there are signs that generative AI could play a more disruptive role in information warfare around the November contest.
Some experts have long predicted that this year’s presidential race would mark the first “AI election,” given that it’s the biggest political event in the US since generative AI output reached new heights of realism in the last couple years.
So far, its role has been overshadowed by unprecedented turns of events in the race. But some recent developments hinted at AI’s reality-warping hazards:
- A recent report from the Microsoft Threat Analysis Center found that Iran-backed groups have laid the groundwork to stoke political divisions around the election further. Those efforts include setting up fake news sites and targeting the email accounts of campaign advisors. The report found evidence that these groups tapped generative AI to create article titles and SEO keywords, as well as to paraphrase real news sites. The Iranian operations build on similar activity the report also observed from Russian and Chinese actors.
- In a testament to how the mere existence of AI might be used to shake trust in visual media, former President Donald Trump falsely claimed that photos of crowds at Vice President Kamala Harris’s Detroit rally had been deepfaked: “She ‘A.I.’d’ it,” Trump wrote on his Truth Social platform. Harris’s campaign denied the claim, and journalists present, media photos and footage, and an expert in digital forensics and misinformation interviewed by the Associated Press confirmed the image was real.
- Five secretaries of state penned an open letter to Elon Musk earlier this month urging him to fix X’s Grok chatbot after it was found to be spreading misinformation about election ballot deadlines, the Washington Post first reported. Grok falsely claimed that Harris had missed the deadline to appear on the presidential ballot in nine states, according to the letter.
The news items come as some tech companies like Meta and Adobe have put in place restrictions aimed at protecting against election-related deepfakes. The Biden administration has also been rolling out efforts to bolster election information security.
As the election season enters what promises to be a hectic final few months, more AI threats could put these protections to the test.