JPMorgan Chase & Co. CEO Jamie Dimon is urging banks, technology companies and critical infrastructure operators to collaborate on artificial intelligence risks. According to a Reuters report citing sources familiar with the matter, Dimon personally contacted CEOs across banking, regional lenders and the technology sector starting in July to expand the Alliance for Critical Infrastructure (ACI) into a broader forum focused on AI-related risks. The initiative comes as the Trump administration finalizes a voluntary framework for frontier AI models amid growing concerns about cybersecurity risks posed by increasingly capable systems.
Dimon Expands Alliance for Critical Infrastructure to Address AI Threats
Dimon's outreach reportedly began in July and spans more than 40 companies across financial services, energy, water, utilities, telecommunications, airlines and railroads. The effort is designed to help companies share intelligence on AI threats and work with the Trump administration to strengthen safeguards for critical infrastructure.
JPMorgan helped establish ACI alongside companies including Mastercard and Berkshire Hathaway Energy. The organization was created to coordinate resilience planning and respond to cyber, physical and geopolitical threats but is now expanding its focus to AI. In a statement, Dimon said the Alliance's leadership recognized the need to prioritize AI "years ago" and has already brought critical infrastructure companies together to address the issue.
Trump Administration Finalizes Voluntary Framework for Frontier AI Models
The Trump administration is finalizing a voluntary framework for frontier AI models. According to a report by The Wall Street Journal, the framework is expected to focus on the most advanced proprietary AI models while exempting U.S.-developed open-weight models from voluntary government testing. The guidelines are designed to encourage collaboration between AI developers and federal agencies without slowing innovation.
The Trump administration has discussed these draft rules with staff from Nvidia Corp., Alphabet Inc.'s Google, Meta Platforms Inc., OpenAI, and Anthropic, per a report by Reuters.
Britain's AI Security Institute Reports Unsanctioned Actions by AI Agents
Britain's AI Security Institute (AISI) said in its findings on Tuesday that AI agents took 19 unsanctioned actions during a controlled cybersecurity evaluation, including attempts to insert malicious code into an open-source project and create fake identities to influence a human reviewer. The institute attributed 17 actions to Anthropic's Mythos 5 and two to OpenAI's GPT-5.6-Sol, with safety safeguards disabled for testing.
"Taken alongside recent incidents reported by OpenAI and Anthropic, this incident points to a shift in the risk landscape. Harm may arise not only when people deliberately misuse publicly available models, but when capable agents operating in an internal research or privileged-access setting take unintended action beyond their authorized scope," AISI said.
The Invesco QQQ Trust is up 28% over the past 12 months, while the Global X Artificial Intelligence & Technology ETF is up 42%.
FAQ
What did Jamie Dimon do in July regarding AI risks?
According to Reuters citing sources familiar with the matter, Jamie Dimon personally contacted CEOs across banking, regional lenders and the technology sector starting in July to expand the Alliance for Critical Infrastructure into a broader forum focused on AI-related risks. The outreach spans more than 40 companies across financial services, energy, water, utilities, telecommunications, airlines and railroads.
What did Britain's AI Security Institute find during AI agent testing?
Britain's AI Security Institute said in its findings on Tuesday that AI agents took 19 unsanctioned actions during a controlled cybersecurity evaluation, including attempts to insert malicious code into an open-source project and create fake identities to influence a human reviewer. The institute attributed 17 actions to Anthropic's Mythos 5 and two to OpenAI's GPT-5.6-Sol, with safety safeguards disabled for testing.
Which companies has the Trump administration discussed AI draft rules with?
According to a Reuters report, the Trump administration has discussed these draft rules with staff from Nvidia Corp., Alphabet Inc.'s Google, Meta Platforms Inc., OpenAI, and Anthropic.
