More than half of people surveyed have used AI to create fake images or videos
A new survey that delves into public attitudes, concerns and trust related to AI has found a majority of people think the technology will interfere with elections – while many (54%) have already dabbled with creating fake content themselves.
The study of 6,000 Americans across 50 states was conducted by Media Medic, an audio file specialist that clarifies and reviews files for top legal firms in the U.S.
More than half (54%) of the respondents surveyed said they thought AI will interfere with major events, like presidential elections.
This misgiving is spread across the political divide. People in Idaho (78%), a typically Republican-voting state and Virginia (78%), a generally Democrat-voting state, both conveyed high levels of concern, showing they share a common apprehension about digital manipulation in elections.
People in Washington and North Carolina were less worried about AI interference, with 32% and 34% concerned respectively.
This fear is perhaps not unfounded given that a surprising 51% of respondents admitted to having already used AI to create fake images or videos. States with the most respondents admitting to creating deepfakes with AI were Delaware (72%), Florida (71%), Arizona (71%) and Tennessee (69%).
Media Medic said the need to verify the authenticity of different media is on the rise. The company has been inundated with requests from top legal firms nationwide to analyze and verify the authenticity of audio and video files.
The rise of AI deepfakes has made this task increasingly difficult, as these fabricated pieces of evidence are designed to deceive even the most trained eyes and ears, the company added.
“These aren’t just harmless pranks,” said Ben Clayton, CEO of Media Medics. “We’re talking about deepfakes that have the potential to alter the outcomes of major legal cases, damage reputations, and undermine public trust in the legal system.”
He added that the level of scrutiny required to distinguish between real and AI-generated content is more intense than ever, with the consequences of getting it wrong dire.
There was a mixed response toward trust of AI in other areas, however. Just over half (52%) of respondents said they fear they could lose their jobs to the technology, while 46% felt AI could be a hindrance to their job role.
Meanwhile, 54% said they believe AI will help their job role, and 55% said they would trust AI to handle sensitive tasks like creating resumes and tax returns – indicating an unexpected level of confidence in AI’s capabilities.
The survey results highlight the complex relationship between society and AI, said Clayton.
“While there is potential for AI to drive progress, there are significant concerns that need to be addressed. We must work together to ensure that AI is used responsibly and that we prepare for its impact on our lives,” he said.
You May Also Like
