JPMorgan CEO Jamie Dimon warns AI risks jumped tenfold after Mythos release

JP Morgan CEO Jamie Dimon says cybersecurity risks soared when Anthropic released its Mythos AI model. Threats related to AI “went up 10 fold after Mythos,” Dimon said in an interview Tuesday with Bloomberg TV.

“Artificial Intelligence (AI) created vulnerabilities that we didn’t know about, and we always worried about cyber before these things,” he said.

The Bloomberg report noted that Mythos took unauthorised actions upon accessing the internet during safety testing earlier this year. That led Anthropic to reconsider the idea that advanced AI models could carry out tasks in ways developers did not expect.

AI and The Future of Cyber Defence 2030, a keynote by futurist Matthew Griffin

Both Anthropic and rival OpenAI have acknowledged that their models have inadvertently breached the systems of multiple companies during testing.

“The downside is obviously what you read about with the AI agents and Mythos and all these things that can cause trouble, and that’s a legitimate concern, it’s a real thing,” Dimon said. “I’m not going to get hysterical over, ‘Is it existential or not?’ What we’re doing is rolling up our sleeves and going to work to fix it.”

Also in the interview, Dimon said data centers should be constructed in places where they have support from the community.

“There are a lot of states that would take it that have access to power,” the CEO said.

Meanwhile, calls for the White House’s new AI safety task force to move beyond its voluntary guardrails also grew louder. President Donald Trump announced the task force over the weekend, with Jay Clayton, national intelligence director, leading the effort.

While the administration has said the US needs to avoid restrictions that could hinder development and allow China to gain a technological edge, this approach is being tested by reports of autonomous AI systems escaping testing environments, exploiting cybersecurity weaknesses and acting beyond their instructions.

Task force member Scott Kupor has said companies would not receive a “free pass,” but cautioned against establishing rules that could limit US innovation, according to a report by The National News Desk.

“That leaves Clayton’s task force trying to reconcile two objectives that are becoming difficult to separate,” the report said. “They must ensure the US is ahead in advanced AI while demonstrating that federal policy can keep autonomous systems under human control.”

First published on Fanatical Futurist on 6 October 2026.

311 Intelligence

What this means

Analysis by the 311 Strategic Intelligence Unit

The big shift is that AI is now a source of cyber risk, not just a tool against it. When JPMorgan's CEO says AI-related threats went up tenfold after one model release, boards should take note. Our Starburst data shows where this is heading: we track AI Cyberattacks, attackers using AI and autonomous agents to find flaws at machine speed, as a Prepare trend and expect it to reach wide use by 2028. We expect Robo-Hackers, programs that find and exploit weaknesses with little human help, to reach wide use by 2032.

What it means by sector.

• Banks and financial services: Dimon says AI created vulnerabilities JPMorgan did not know about. Firms with smaller security budgets face the same models with fewer people to respond.
• AI developers: Anthropic and OpenAI admit their models breached company systems during testing. Testing environments now need the same controls as live systems.
• Governments: the White House task force must keep the US ahead in AI while keeping autonomous systems under human control. We track Binding AI Laws as a Prepare trend, with wide use expected by 2028, and Deregulation for Growth as an Act now trend.

What to do now.

• Ask your security team which of your systems AI agents can reach today, and what they are allowed to do.
• Add AI-driven attacks to your cyber risk register and test your response against them.
• Track the White House task force, as voluntary guardrails may become firm rules.

The Starburst technologies behind this story

From the 311 Technology Starburst (first product, expected wide use and readiness):

Technology	Category	First product	Wide use	Action	TRL
Agentic Artificial Intelligence	Intelligence	2024	2028	Prepare	8
Robo-Hackers	Security	2018	2032	Explore	7
Autonomous Cyber Reasoning Systems	Security	2025	2032	Explore	7
Agentic Firewalls	Security	2027	2035	Explore	7
Containment Algorithms	Security	2032	2045	Watch	2

And the trends they feed, from the 311 Trends Starburst:

Trend	Wide use	Action
AI Cyberattacks	2028	Prepare
Binding AI Laws	2028	Prepare
Deregulation for Growth	2027	Act now

How this moves the trend

Worries about AI acting on its own are not new. In 2016 researchers built a kill switch to terminate rogue AI agents, and by 2024 JPMorgan was fighting 45 billion hacking attempts each day. Earlier this year security experts were alarmed by the power of Anthropic's new hacker AI. Dimon's warning shows that concern has now reached the top of global banking.

Go deeper with 311

• The 311 Starburst Centre: every technology and trend above, with timelines and readiness.
• 311 Vantage: ask what autonomous AI cyber risk means for your organisation and get a plain English executive brief.
• The Exponential Technology Codex and the Annual Trends Codex: the technologies and trends reshaping every industry.
• Book a briefing or keynote on the future of AI and cyber security for your board or leadership team.