Chase Focuses on Reliability and Security to Develop Its AI

Key Takeaways

1Chase emphasizes the reliability of AI agents through safeguards and assessments

2Data security and regulatory compliance are top priorities

3An internal platform, LLM Suite, was launched in the summer of 2024 for all employees

Why it matters : Chase's strategy illustrates how a major bank is adapting its practices and skills in the age of agentic AI.

At Chase, AI relies on safeguards, strict data protection rules, and an internal platform launched in the summer of 2024. Its CIO, Gill Haus, prioritizes engineers who can assess the effectiveness of agents rather than write code, while all employees are trained in these tools.

Reliability of Agents and Demand for Scale at Chase

Gill Haus describes agent technologies as systems where models serve as reasoning engines by predicting the next statistically probable action. He emphasizes that this probabilistic nature can lead to unexpected consequences. To establish trust and enable large-scale use, he believes that safeguards, assessments, and outcomes are necessary. He specifies that the underlying architecture must be robust so that services can be deployed safely and productively.

Data Protection and Compliance as Prerequisites

The CIO places the security of AI product delivery and data protection at the forefront. He reminds that Chase, a bank operating in a highly regulated sector, applies high standards to emerging technologies. He insists on consent and confidentiality, specifying that the use of data must remain transparent to customers and compliant with privacy and regulatory requirements.

Le brief IA que lisent les pros

Le brief IA que les pros lisent chaque soir

Les 7 actus IA du jour, décryptées en 5 min. Gratuit.

Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.

Choisissez votre rythme

Gratuit · Pas de spam · Désabonnement en 1 clic

An Internal LLM Suite for Designing Products and Services

Chase launched LLM Suite, an internal agentic platform in the summer of 2024, which provides secure access to large language models, whether cutting-edge or open source. Employees can use it to ask questions, review documents, and generate specifications, and they utilize it to develop new products and services. For example, machine learning helps understand customer intent and quickly direct them to the right person during calls. The entire organization is trained in the effective use of these tools.

Expected Skills: Assessing Agents Rather than Coding

According to Gill Haus, artificial intelligence is already changing the way teams work, including outside the IT sector, and now, almost anyone can design an application thanks to AI-assisted programming tools. He expects engineering specialists to be able to evaluate in which situations agents operate effectively and reliably. He clarifies that he does not recruit engineers to write code but favors those who can assess the functioning of agents. For him, creating value with AI is not limited to choosing a trendy model or service. He serves as the Chief Information Officer at Chase, the retail bank of JPMorgan Chase & Co.