backOn June 25, 2024, Google’s DeepMind, led by Demis Hassabis, released its first comprehensive study on the malicious use of artificial intelligence. The report found that deepfakes—fabricated images, videos, and audio imitating real individuals—have become the most prevalent form of AI misuse, surpassing cyberattacks and accounting for 27% of documented incidents between January 2023 and March 2024.
The research indicates that manipulating public opinion is the primary motivation behind the malicious use of generative AI, heightening concerns about the potential influence of synthetic media on elections worldwide in 2024. Conducted in collaboration with Google’s Jigsaw research unit, the study also revealed that AI tools are being exploited for profit, including services that create deepfakes or mass-produce false news. DeepMind emphasized that most recorded incidents involved easily accessible tools requiring minimal technical expertise, enabling a wider range of malicious actors to misuse generative AI.
share
