Dive Brief:
- Enterprises were quick to jump on the generative AI bandwagon, and they might be even faster to adopt AI agents, according to a PagerDuty report published this week. The company surveyed 1,000 IT and business executives across the U.S., U.K., Australia and Japan.
- More than 3 in 5 decision-makers expect agents will yield a more than 100% ROI with the average anticipated return landing around 171% of their investment. U.S.-based companies are even more optimistic, putting that number closer to 192%.
- Nearly 45% of leaders expect agentic AI to have a greater impact than generative AI. More than 9 in 10 of those surveyed believe they will adopt agents at a faster rate than they did generative AI.
Dive Insight:
Enterprises had a hard time wrangling returns on generative AI investments last year. The disappointment was even more stark when considering where enterprises had hoped they’d be. But those hardships might have smoothed the path for agentic AI adoption, according to the PagerDuty report.
“Rushing in too quickly, overspending on tools, and lacking the right infrastructure left some organizations struggling to see real ROI,” PagerDuty said in a blog post published Tuesday. “Now, as the next phase of AI arrives, businesses are applying these hard-won lessons to ensure they’re using AI effectively.”
The majority of businesses looking to adopt agentic AI started first with generative AI, according to PagerDuty’s survey. Companies that have gone steps beyond adoption and adapted governance frameworks and responsible practices could find it easier to adjust for agentic AI compared with organizations starting from scratch, analysts have told CIO Dive.
Vendors are also trying to accelerate adoption by stepping in to mitigate technical barriers to agent implementation with pre-built use cases, tailored solutions and low-code development tools.
Despite enthusiasm, IT leaders still have concerns. Most decision-makers point to data security and privacy, legacy integration challenges and a lack of employee understanding, according to a SnapLogic survey.
Enterprises can exacerbate risks without the right foundation. IT and business leaders said the top risks from implementing agentic AI are security vulnerabilities and AI-powered cyber attacks, the PagerDuty report found. Most companies are planning to upskill employees via enterprisewide seminars, external courses or official office hours and internal mentorship programs to ensure agents are used in the intended fashion.
Technology leaders across industries, from Walmart to Expedia and SharkNinja, have emphasized the need to balance caution and innovation. Where exactly to draw the line depends on an organization’s risk appetite and ability to implement guardrails.