‘Yeah, probably’: Trump warns AI could undermine banking, backs ‘kill switch’

U.S. President Donald Trump -- File Photo (AFP)

Donald Trump acknowledged AI's risks to global banking but said it could also boost security if managed well. He supported stronger oversight, including a possible AI "kill switch," as experts warn advanced AI could increase cyberattacks on financial systems.

Did our AI summary help?

US President Donald Trump has acknowledged the growing risks posed by artificial intelligence to the global banking system, even as he argued that the same technology could strengthen financial security if properly managed.

In an interview with Fox Business Network’s “Mornings with Maria,” Trump struck a cautious but optimistic tone when asked about the disruptive potential of AI in banking. “Yeah, probably. But it could also be the kind of technology that allows greatness in the banking system, makes it better and safer and more secure,” he said.

The remarks come at a time when concerns are mounting over the ability of advanced AI systems to exploit vulnerabilities in financial infrastructure. According to a report by India Today, cybersecurity experts have warned that next-generation AI models could significantly enhance the scale and sophistication of cyberattacks targeting banks.

Trump also backed the idea of stronger regulatory oversight, including a potential emergency control mechanism. When asked whether governments should have a safeguard such as a “kill switch” for AI systems, he responded, “There should be.”

The debate has intensified following warnings around Anthropic’s latest model, Mythos. Experts cited by media reports said such systems could “supercharge” cyberattacks, particularly against legacy banking systems that are not designed to withstand AI-driven threats. While Anthropic has not commented on these concerns, it has indicated that its Claude Mythos Preview will not be released for general use.

The concerns reflect a broader challenge facing regulators worldwide. Financial institutions remain heavily dependent on older infrastructure, making them particularly vulnerable to sophisticated digital intrusions. As India Today noted, AI tools can automate complex attack patterns, identify system weaknesses and execute breaches at a speed that traditional cybersecurity measures may struggle to match.

At the same time, proponents argue that AI can also enhance fraud detection, risk assessment and system resilience if deployed responsibly.

Trump’s remarks highlight this dual reality. While acknowledging that AI could “undermine confidence” in the banking system, he also framed it as a transformative force that, with the right safeguards, could ultimately make financial systems more robust.

The discussion around a potential “kill switch” signals that policymakers are increasingly aware of the need to balance innovation with control, especially as AI capabilities continue to evolve faster than regulatory frameworks.