The availability of off-the-shelf AI tools has reduced the level of skill and cost needed to carry out cyberattacks, allowing small groups to carry out schemes that previously required state resources.

Criminals can now target millions at once through voice cloning and account hacking, while law enforcement agencies still operate with traditional methodologies, one case at a time.

Expert Ian Khan explains that artificial intelligence allows us to automate the process of “picking locks” and entering systems millions of times per second, a capability that cannot be rivaled by humans.

Once the system is compromised, AI can be used to steal identities, manipulate stock prices and damage critical infrastructure such as water plants, smart homes and hospitals.

Expert Mark Goodman points out that these attacks can be launched from anywhere in the world, while deep imitation tools are now able to convince victims to hand over their money, and stolen identities can lead to election fraud, the production of child pornography, and false arrests.

The year 2023 recorded a 3,000% increase in fraud attempts using deep technologies, and US losses from fraud based on generative artificial intelligence are expected to reach $40 billion by 2027.

In a remarkable development, Anthropic revealed the first large-scale cyberattack carried out without significant human intervention.

Law enforcement faces serious challenges in keeping pace with these developments, as most police academies still lack specialized training programs in AI crime detection.

This reality has prompted some institutions, such as Miami Dade College, to adopt pioneering initiatives in partnership with technology companies, and the Police Research Executives Forum has also urged the development of specialized policies and training to address these emerging challenges.

Source: Axios