DragonImages/ iStock / Getty Images Plus via Getty ImagesZDNET’s key takeaways
- Agentic AI shifts professionals’ responsibilities.
- Embed agents deep into operational processes.
- Secure services and reliable outputs are crucial.
Evidence suggests that delivering value from AI is hard. With 91% of professionals saying their firm still falls short on AI, business leaders and professionals have lots of work to do to ensure emerging tech explorations become useful production services.
Also: Why replacing staff with AI backfires: 5 ways smart leaders generate real value instead
Gill Haus, chief information officer at Chase, the consumer banking arm of JPMorgan Chase & Co, recognized that creating value in the age of AI is about much more than deciding to implement a hyped-up model or service.
“Saying that you’re going to be efficient by putting a target in place is not the same thing as reimagining your processes through an agentic lens,” he told ZDNET.
Haus said it’s crucial to understand that roles and responsibilities will shift, and AI is already transforming the work that Chase staff fulfill, both in the IT department and beyond.
“We are seeing that our traditional engineer, who was writing code in the past, can do more than just write code,” he said. “And a product leader in the business, who traditionally was creating stories, can now write more code.”
As agents become part of the internal operating fabric, the resulting workplace will blend AI and humans.
Haus said it’s already possible to see this transformation at Chase, particularly in coding, and the people who fulfill engineering roles. This impact shows up in ways that would have seemed surprising only six months ago.
Also: ‘Specialists aren’t required’ anymore: How to stay valuable in an AI agent workplace today
“I don’t really hire engineers to write code. Now, I know that sounds weird because that’s what I should hire engineers to do,” he said.
“But today, I hire engineers to know what code to write. There’s a big difference. Until six months ago, you had to write the code. But now you don’t have to because an agent can help you. This capability means a lot of the tedium that got in the way of what engineers really like to do goes away.”
So, with a tight focus on outputs and not just inputs, how does Haus ensure Chase has the right approach to AI? The answer centers on three core fundamentals: creating seamless services, delivering secure products, and enabling reliable outputs.
1. Create seamless services
Haus said his organization thinks deeply about AI’s role in the operating model, with all staff trained to use these tools effectively.
A crucial element of this approach is LLM Suite, the firm’s internal agentic platform that staff can use to ask questions, review documents, and create specifications.
Also: 3 surveys deliver the same uncomfortable truth about adopting agentic AI
LLM Suite was released in summer 2024 and provides access to large language models, both frontier and open-source technologies, in a secure environment.
“It’s a platform that anyone in the organization can use,” he said. “We’re looking at the end-to-end process and using the technology to remove the things that got in our way and to unleash our teams across the organization.”
Haus explained how employees use the agentic platform and its models to create new products and services.
“If there is an experience that we have and we can use any technology to make it better for our customers, more personalized, we are doing that.”
Also: Businesses must reinvent their processes and workforce to scale agentic AI adoption
He gave the example of using emerging technology to improve the experience of a customer who calls the bank.
“We use machine learning to understand intent so we can get you to the right person quickly,” he said. “When you are on the phone with us, we use similar technology to figure out what you may want to do.”
Haus said the key to success is embedding agents and algorithms deep into the operational process, whether that’s a staff member using an internal interface or customers using a web service or mobile app.
Emerging technology is available to Chase employees who work with the systems every day, but its inner workings must be seamless to people who use the bank’s services.
“The real magic is making things easier for a customer, so they don’t even realize that the experience has been made better,” he said. “It just works for them.”
2. Deliver secure products
While great AI-enabled customer services are crucial, those products must be delivered securely, with the highest possible priority placed on data protection.
As a bank operating in a highly governed sector, Haus said Chase has a high bar for emerging technologies: “We are thoughtful, and we follow the practices we’ve had in the past, where we can move quickly but responsibly.”
Also: The best and worst AI for your privacy, ranked – and how each handles your data
Data protection is an absolute must, as is consent and privacy.
“If you use data for something, we ensure our customers know,” he said. “We would never use the data in an insecure way or against any of our privacy, compliance, or regulatory guidelines. Whether it’s traditional machine-learning models or agentic technologies, those practices are critical.”
Security isn’t just a customer-facing consideration. Haus said AI also assists Chase behind the scenes, particularly in developing software quickly and effectively.
Haus said the AI-enabled tools professionals use help them identify bugs, prevent bad actors, and deliver better customer experiences.
Also: How to keep your conversations with ChatGPT, Gemini, Copilot, or Claude private
For other professionals and business leaders, the watchword for secure software development in the AI age is proactivity.
“The thing that I believe is incredibly important is continuing to make sure, outside of what we do to make AI work well, that we’re focused always on our perimeter, ensuring that we are keeping our software updated, and that we are working on addressing vulnerabilities in a timely fashion,” he said.
“This proactive approach ensures that, as the world around us shifts, we can protect our customers.”
3. Enable reliable outputs
Agentic technologies use models as a reasoning engine to predict the most statistically likely next action. Unfortunately, agents, just like their human counterparts, are probabilistic — and Haus said this nature can have unintended consequences.
“You won’t always get the same answer,” he said. “This outcome matters less when you’re looking for a recipe. If you end up with a different type of mayonnaise, it might be an issue. But when you want to do something with your finances, the output needs to be correct.”
Haus said business leaders must ensure guardrails, evaluations, and results build trust and create confidence that agentic technologies can be used at scale.
“In many cases, the solution is to put the human in the middle, and so there are a lot of deployments at Chase where that’s the case.”
Also: AI agents are your new colleagues – how to get the best results
He returned to software development and his assertion that Chase doesn’t hire engineers to write code.
In an age of agents, almost anyone — IT professional or not — can build an app using AI-enabled coding tools.
Haus wants talented engineering experts who can tell when agents are working effectively and truthfully, and he advised other business leaders to do the same.
“Thinking through and providing clarity on what must be true from an outcome when you are building software — from scale to the components that must be used to how it should be architected — are the hard parts,” he said.
Also: 45% of professionals use shadow AI tools – here’s how to manage the risks
Chase wants its people to use agentic AI, but the underlying architecture must be strong so services can be scaled safely and productively.
“What are the patterns? How do we make sure those patterns can be followed so that, when our people build software, they’re building the outcome the way they want, not just getting something the model is spinning out that isn’t thought through holistically?” he said.
“That’s the piece that I think is a fundamental that we need to capture. And you can do that in a variety of different ways in an organization, more than just using your talented engineers to write code.”
