Topics: Donald Trump, Technology, Artificial Intelligence, US News, World News, News
Donald Trump has spoken to banking leaders in the US about the threat posed to financial systems by artificial intelligence.
Trump was joined by Treasury Secretary Scott Bessent, as well as Jerome Powell, the Chairman of the Federal Reserve, in a session at the headquarters for the Treasury in Washington DC on April.
According to reporting in Bloomberg, the meeting with senior bankers in the US was called specifically for banks deemed to be of systematic importance both in the US, and around the world.
Among the attendees were Brian Moynihan of Bank of America, David Solomon of Goldman Sachs, Jane Fraser of Citigroup, Charlie Scharf of Wells Fargo, and Ted Pick of Morgan Stanley, according to a report in the Daily Mail.
Advert
The meeting concerned Anthropic, and a new AI model that the business has produced called Claude Mythos.
Anthropic has warned about the capabilities of this technology, advising that it has the capability of hacking into secure networks.
The company made a blog post about their new model, writing: "AI models have reached a level of coding capability where they can surpass all but the most skilled humans at finding and exploiting software vulnerabilities."
It warned: "The fallout – for economies, public safety, and national security – could be severe."
One case study saw Calude Mythos identify a weakness in a piece of software called OpenBSD which had been there for 27 years.
OpenBSD has a reputation for its security, and human technicians had not previously spotted the problem.
Once identified, this allowed someone to crash computers remotely, simply by connecting.
In another incident, the AI model was able to combine together a number of weaknesses in Linux kernel.
This is a type of software which operates most computer servers around the globe, and Anthropic has warned that this could allow a hacker to 'escalate from ordinary user access to complete control of the machine'.
Dr Roman Yampolskiy is an AI safety researcher at the University of Louisville, and said that the existence of this AI model at all is a problem.
Speaking to the New York Post, Dr Yampolskiy said: "Ideally, I would love to see this not developed in the first place. And it's not like they're going to stop.
"That's exactly what we expect from those models – they're going to become better at developing hacking tools, biological weapons, chemical weapons, novel weapons we can't even envision."
The system is now a cause for widespread concern in the global financial industry due to the potential risks it opens up to economic stability.
