JPMorgan rolls out AI tools across investment banking as Dimon signals hiring shift

TL;DR:

• Paul Uren, JPMorgan’s Asia Pacific head of investment banking, has confirmed the bank is implementing AI tools globally across its IB business, saying tools “enable us to access more information and quickly synthesize it with our internal systems”.
• On the same day, CEO Jamie Dimon told Bloomberg that JPMorgan will “be hiring more AI people and fewer bankers in certain categories”, adding that AI “will reduce our jobs down the road” but a c.10% annual attrition rate gives the bank flexibility to manage the transition through retraining and redeployment rather than large layoffs.
• JPMorgan is among a small group of organisations Anthropic has admitted to its “Project Glasswing” controlled access programme for the Mythos cybersecurity model, alongside Goldman Sachs, Citigroup, Bank of America and Morgan Stanley.

The Asia rollout and the Dimon Bloomberg interview together amount to the clearest UK-relevant statement of intent yet from a global bank on AI’s role in investment-banking labour. Uren’s “early phase” language sets expectations against a multi-year rollout; Dimon’s “more AI people, fewer bankers” framing sets the directional commitment. The interesting question is what this means for the City of London’s IB headcount, where JPMorgan employs roughly 19,000 people across its UK operations.

Three banks in 72 hours: a coordinated pivot or a coordinated message?

The week’s AI-and-jobs banking news has been remarkably synchronised. Standard Chartered announced 7,000 job cuts on Tuesday with AI substitution central to the plan, drawing supervisory questions from MAS and HKMA (see [our coverage]). HSBC CEO Georges Elhedery warned staff that AI would “destroy and create” jobs. Now JPMorgan. Whether this is genuinely a coordinated industry pivot or a coordinated industry message is the harder read. Dimon’s specific framing — managing the transition through normal attrition rather than layoffs — is notably gentler than Winters’s “lower-value human capital” wording at StanChart, and looks designed to inoculate JPMorgan against the supervisory response StanChart has just attracted.

The Project Glasswing detail matters separately. JPMorgan’s permitted use of Anthropic’s Mythos cybersecurity model means the bank is operationalising AI not just for productivity but for live security tooling — exactly the use case Anthropic has warned could supercharge complex cyberattacks if it leaks. For UK financial-services security and risk leaders, the practical question is whether the audit trail on Glasswing-class deployments will satisfy the FCA’s parallel push for evidence-based AI assurance.

Looking forward

Watch for analyst questions on JPMorgan’s H2 2026 calls about UK IB headcount trajectory specifically, given that the Dimon framing applies “in certain categories” rather than across the board. Other UK-listed and UK-active wholesale banks will be under increasing pressure to articulate AI-driven workforce plans in similar language — gentler than Winters, more concrete than HSBC’s. For UK fintechs offering banker-augmentation tools, the Uren quote — “streamlines the preparation of content and materials, as well as helping bankers engage with more clients more efficiently” — is the live procurement brief.