Listen to this article. This audio was generated by AI.
Artificial intelligence will be playing an even bigger role in enterprise business processes sooner rather than later. The evolution of agentic AI, or AI agents acting on goals and decisions, will take business process automation to the next level.
That's according to analysts speaking during the Forrester Research Technology and Innovation Summit in Austin, Texas, on Tuesday. Akin to self-driving cars that have been slowly modified to transition driving from humans to machinery, agentic AI could eventually automate multiple steps within complex processes currently performed by humans, such as receiving and processing insurance claims.
"There are dozens of steps in that process. Some of them are automated; some of them are not," Forrester Research analyst Mike Gualtieri said during a panel session at the summit. "There's a claims processing person who's in charge of that claims and that process. Agentic AI is where AI replaces that person and becomes in charge of that process. It's full-on automation of that process, and it's not far-fetched at all for tens of thousands of business processes."
Agentic AI, especially in the early days of development, will mean putting whole business processes on automatic pilot but should also include having someone monitor the AI system, Gualtieri said. He noted that self-driving cars are an apt comparison to agentic AI. While self-driving cars are on the road, they aren't left unmonitored.
"Agentic is about automating the entire process," Gualtieri said. "To do that, you have to still automate the steps in the process."
To make agentic AI work, Gualtieri said it's not just about automating and orchestrating a business process. Agentic AI will require collaboration proficiency, either systems working with people in different departments or with other AI models.
While business process automation is not new, it can be a complex undertaking to include rules and codes within automated process descriptions. Forrester Research analyst Brian Hopkins said more sophisticated neural networks are enabling technology like agentic AI to break down complicated tasks, use different tools to accomplish those tasks and communicate with other AI models.
It's full-on automation of that process, and it's not far-fetched at all for tens of thousands of business processes.
Mike GualtieriAnalyst, Forrester Research
"To me, it's about the tools," Hopkins said during the panel discussion with Gualtieri. "These agents can use tools like we humans at one point learned to use tools, and that gave us a huge step in our ability to advance. The same thing is happening here."
For enterprises to automate processes using agentic AI, it will require a team beyond just AI experts, Gualtieri said. It will involve company leaders who understand different facets of business operations as a whole and change management.
Safety planning will also be crucial for agentic AI, Gualtieri said. Safety guardrails could include businesses training their own AI models with their own data. He said it could also involve creating a prescribed process for agentic AI, outlining what it can and cannot do.
"The first aspect of that is not just having some AI model spitting out [information] that has no transparency but having an agentic AI model that will follow a prescribed process you've designed," Gualtieri said.
Makenzie Holland is a senior news writer covering big tech and federal regulation. Prior to joining TechTarget Editorial, she was a general assignment reporter for the Wilmington StarNews and a crime and education reporter at the Wabash Plain Dealer.
