Dive Brief:
- While the majority of workers are open to using AI agents, many have concerns about the technology’s reliability, accuracy and transparency, a Pegasystems report published last month found.
- One-third of the labor force worries about the quality of AI agents, according to a survey of more than 2,100 U.S. and U.K.-based workers conducted in partnership with YouGov. Other respondents questioned the technology’s accuracy and lack of emotional intelligence.
- Around 2 in 5 respondents said they feel uncomfortable submitting AI-generated work, and more than one-third of workers said AI-produced work is subpar compared to their own.
Dive Insight:
AI agents are on enterprise to-do lists this year, but most organizations remain in the prep phase.
On the technical side of implementation, most CIOs admit to needing tech stack upgrades before pursuing AI agents. Vendors have stepped in to lower barriers to adoption, but there are still questions about standardization and governance.
Ensuring the security of the growing attack surface presents another enterprise hurdle.
Agents are likely to become prime targets for threat actors, underlining the need to strengthen security practices ahead of introducing agents to workflows, according to Gartner. In the next three years, the firm predicts AI agent abuse to be responsible for 1 in 4 enterprise breaches.
Technology leaders also have to contend with worker sentiment as adoption unfolds. More than 2 in 5 workers said enhanced accuracy and reliability would ease concerns, followed by better training and increased transparency, according to the Pegasystems report.
While still rare, some enterprises have started to deploy agents. Accenture has around 600 marketing pros using agents to craft campaigns. Toyota Motor Corporation has around nine agents helping share and store knowledge.
