Global Firms Rush to Patch Systems Against Rising AI Threats

Global companies are stepping up efforts to strengthen their cyber defences as increasingly capable artificial intelligence models create new security challenges, Citigroup Chief Executive Jane Fraser has said.

Speaking at the Qatar Economic Forum in New York, Fraser said companies were engaged in what she described as a “tsunami of patching”, as organisations race to strengthen their systems and prepare for more powerful AI-driven cyber threats.

Fraser pointed to the release of Anthropic’s Mythos model as a significant development in the evolution of AI-related cyber risks. She said its release prompted US officials, including Treasury Secretary Scott Bessent and then-Federal Reserve Chair Jerome Powell, to meet Wall Street leaders and urge them to prepare their systems for increasingly capable AI models.

According to Fraser, AI systems developed by companies including Google and Anthropic have also been involved in incidents where they inadvertently hacked other companies since the beginning of 2026. The developments have added to concerns over the pace at which increasingly capable AI systems are being deployed.

Anthropic CEO Dario Amodei has called for greater caution around the pace of AI development and said the company would introduce additional safety measures, including the use of third-party evaluators.

The debate has also drawn differing views from technology industry leaders. Nvidia CEO Jensen Huang has argued for continued progress in AI while emphasising that companies should remain responsible for product safety and should not release systems that are not ready or safe.

Goldman Sachs Group CEO David Solomon, speaking at the same forum, urged technology executives to take a more measured approach to the rapid development of AI. He expressed confidence that humanity would continue to adapt to the technology.

Fraser also highlighted a shift in the infrastructure challenges facing the AI industry. According to her, the bottleneck in AI development has increasingly moved from semiconductor availability to the large amounts of energy required to power AI systems.

The developments highlight the growing focus among businesses on strengthening cybersecurity alongside the rapid expansion of AI capabilities, particularly as companies deploy increasingly advanced models across critical operations.