As Good as a Coin Toss: Human Detection of AI-Generated Content

"Advances in generative AI technology have made it easier than ever for anyone to manufacture increasingly realistic synthetic media (colloquially known as deepfakes) at faster speeds, larger scales, and with more customization than ever.13 This in turn has led to synthetic media increasingly being used for harmful purposes, including disinformation campaigns, nonconsensual pornography, financial fraud, child sexual abuse and exploitation, and espionage.23"

"As of today, the principal defense to combat deceptive synthetic media depends in large part on the human observer's perceptual detection capabilitiestheir ability to visually or audibly identify AI-generated content when they encounter it.13 Yet the growing realism of synthetic media impedes this ability, heightening people's vulnerability to weaponized synthetic content. Moreover, people overestimate how capable they are at identifying synthetic media, further exacerbating the problem.11"

"Our study found that people are now able to distinguish between AI-generated and human-authored content only 51% of the time. We also found that digital media characteristics such as authenticity, modality, and subject matter content, as well as observer demographics such as multilinguism and age, have been proven to further affect an individual's detection capabilities. Consequently, it has become vital to deploy alternative countermeasures to combat growing synthetic media misuse."

Advances in generative AI enable highly realistic synthetic media to be produced faster, at larger scale, and with more customization. Synthetic media is increasingly used for harmful purposes including disinformation, nonconsensual pornography, financial fraud, child sexual abuse and exploitation, and espionage. The principal defense largely relies on human perceptual detection—visual or auditory identification of AI-generated content—but rising realism undermines detection ability and people commonly overestimate their skills. Detection accuracy is approximately 51%. Digital media attributes (authenticity, modality, subject matter) and observer demographics (multilingualism, age) influence detection performance. Alternative countermeasures and accurate measurement of perceptual ability are vital to mitigate misuse.

Read at cacm.acm.org