At a House Panel hearing on Wednesday, policy experts expressed a harsh assessment of the growing criminal use of artificial intelligence.
Rep. Andy Biggs, R-Ariz. The chair of the House Judiciary Subcommittee on Crime and Federal Surveillance began the hearing, stating that “the threat of AI response continues to evolve as bad actors use AI technology in a wide range of criminal enterprises.”
The hearing comes after Congressional Republicans decided they opposed the inclusion of a 10-year moratorium Regarding the regulation of individual states of AI in one big beautiful bill law. Without federal AI regulations, states have already begun regulating new technologies through law. In particular, Texas has restricted its use to censor the use of AI on some websites and requires the right to opt out of AI-driven personal data harvesting.
Andrew Bowne, professor of AI law and policy at George Washington University Law School, explained how emerging technologies associated with the term artificial intelligence allowed criminals to commit crimes on a larger scale.
“We accelerate traditional processes, but we also create whole new ones. If the tasks that AI is used are criminal or harmful, the nature of AI will be a multiplier of the threat,” explained Bowne.
Bowne cited computer vision systems, generative adversarial networks (GANS), and large-scale language learning models (LLM) as three major forms of technological advancements currently used by criminals to exploit their citizens.
Bowne noted that computer vision systems are “used to automate surveillance, identify targets, and harvest personal data from compromised documents to support identity theft and fraud.”
“What allows real-time threat detection for public safety can be reused for individuals with stalks and terrifying mail with calm efficiency,” says Bowne I warned him.
Law school instructors then detailed how criminals can utilize audio and video online and weaponize it to force individuals.
“Gans can produce composite images, videos and audio that are well known in public discourse as 'deepfakes'. These tools allow for the impersonation of both civil servants and civilians,” added Bowne.
They are involved in overwhelming financial institutions with expanded realistic conversations, targeting seniors and vulnerable people targeted with scams, and with thousands of customized loan applications.
It is also used to generate malicious code, allowing individuals without technical backgrounds to access cybercrime.
Ari Redbord, global policy director at TRM Labs, explained that hearings are important as the entire criminal ecosystem is changing, as a software company providing security services and a former US lawyer.
“When the marginal cost of a fraud, phishing campaign, or attempted to fear approaches zero, the complexity increases exponentially,” Redboard warned.
“We're not just looking at the same thing. We see new types of threats that are impossible before AI. We see new types of fraud, hyper-personalized fraud, deepfake horrors, autonomous laundry washing,” said a cybersecurity expert.
Still, Redbord warned against a potential ban on artificial intelligence.
“The solution to AI's criminal abuse is not to ban or suppress technology, but to use it wisely. We need to stay one step ahead of the illegals by leveraging the same innovations they use for their bad,” he said.
