Dario Amodei, CEO and co-founder of Anthropic, left, and Jamie Dimon, CEO of JPMorgan Chase.
Denis Balibouse | Reuters | Samuel Corum | Bloomberg | Getty Images
Anthropic CEO Dario Amodei warned Tuesday that artificial intelligence has created a narrow window for tech firms, governments and banks around the world to fix tens of thousands of software vulnerabilities found in his company’s latest model.
That AI model, Mythos, was previewed last month along with the revelation that it had uncovered decades-old vulnerabilities in key software.
Since AI models from geopolitical adversary China are “perhaps six to 12 months” behind the Anthropic product, there is “about that much time” left to fix these issues, Amodei said.
The comments came during an Anthropic event where Amodei shared the stage JPMorgan Chase CEO Jamie Dimon unveiled a new suite of agents designed to automate financial work.
“The danger is simply a huge increase in vulnerabilities, the number of breaches and the financial damage caused by ransomware in schools, hospitals and not to mention banks,” Amodei said.
Anthropic has limited Mythos to a few partner companies because of concerns about what criminals or enemy nations might do with it. The company’s recent model updates have reverberated across markets, but Mythos has caused the greatest concern among both companies and policymakers.
The scale of potential cyber exploits has increased with each generation of Claude, Amodei said. An earlier version of Anthropic found around 20 vulnerabilities in the Firefox browser. Mythos has found nearly 300, and the total number of all software programs is now in the tens of thousands, he said.
Most of the vulnerabilities Mythos found have not been publicly disclosed because they are still unpatched and “the bad guys will exploit them” if they are identified, Amodei said.
“A better world”
Despite the concerns, both Amodei and Dimon expressed limited optimism.
“This is about a moment of danger where if we respond correctly we can have a better world on the other side, and I think we have begun to take the first steps,” Amodei said. “There are only so many mistakes to find.”
Dimon also said that while cyber fears are justified, cybersecurity risks caused by AI are a “transitional period.”
On the question of regulation, Amodei said that AI oversight should be similar to what is happening in the automotive industry and should strike a balance between consumer safety and industry competitiveness.
“You can’t just start a car company without asking, ‘Are there brakes on this thing?’” he said. “We need to move towards a process that allows the industry to move quickly, that is fair but puts limits on the most serious things.”
The company’s event and setting, featuring Dimon, the finance industry’s best-known speaker, appeared to demonstrate Anthropic’s lead over OpenAI in the enterprise AI market as both companies head toward possible initial public offerings.
Anthropic announced on Tuesday an expansion of its financial services platform, including 10 new AI agents for investment banking and back-office work, as well as integration with Microsoft’s various Office programs. The company also said its latest widely used model – Claude Opus 4.7 – tops benchmarks for financial analysis tasks.
Choose CNBC as your preferred source on Google and never miss a moment from the most trusted name in business news.