Our Editor, Evie, was pleased to sit down with Ryan Loftus, representing J.P. Morgan Chase, at the recently wrapped up Identity Week America conference.
As he excellently put it, a web of vulnerabilities lie in traditional cybersecurity and identity systems which are fundamentally reactive and fragmented and downstream authentication steps have not been designed for an “agentic” AI world.
The “Least Common Denominator” problem has arisen because consumers bank, shop, and share data across multiple platforms, and threat actors can easily exploit weak links across the broader network. A security patch at one institution does not secure the whole ecosystem.
These systemic risks to security include a lack of initial verification, with the most historic user authentication distinguishing humans and growing non-human actors.
This relies on approximation and pattern tracking of transactional, physical or digital behaviours rather than truly verifiable, reusable identities.
How does JP Morgan adopt AI in proactive protection strategy?
In this interview, Loftus provides an insight into extensive corporate strategy of JP Morgan where a user’s identity is verified up front, leverages machine learning models, biometrics, voice recognition, and anomaly detection built over years. Their internal capabilities and scale buffers an ecosystem of threats, but as Loftus says, they couldn’t solve the challenges alone and actively partner with external technology providers to integrate frictionless solutions.
We delve into discussing stakeholder cooperation to set new uniform standards that control bad actors who currently operate without the regulatory barriers. In particular, agentic AI needs definition through standards work. All this and more is discussed in depth between our Editor, Evie and Ryan Loftus.
