This audio is generated automatically. Please let us know if you have comments.
Diving brief:
- Companies move more AI pilots in concert production with reinforced governance, according to the KPMG Tribestly Pulse Survey for the third quarter, published Thursday. The report studied 130 chiefs of suite C and business representing organizations with an annual turnover of $ 1 billion or more in the United States
- Over 2 out of 5 companies actively deploy AI agents, Coming in relation to 11% in the first quarter and 33% in T2. The number of companies exploring and piloting technology has dropped as more projects reach a wider implementation.
- With more agent experience under their belts, companies strengthen railings and surveillance. More than 3 out of 5 leaders said they put humans in the loop due to a lack of confidence, 45% in the last quarter. More managers also limit access to agents to sensitive data, going from 45% to T2 to 63% this quarter.
Diving insight:
Companies have high aspirations for AI agents, but the risks are also high. To alleviate potential dangers, CIOs have resumed their governance manuals.
“The traditional AI was static – you formed it, deployed, monitored it”, ” Bryan McGowan, the chief of the confidence at KPMG USsaid diving by CIO. “But agentic AI systems can perceive, reason, plan and even act independently.”
For the moment, AI agents remain mainly in the It, risk and operations functions, such as quality insurance and fraud prevention, Kpmg find. WalmartFor example, uses AI agents in its technology service to Identify accessibility gaps in code and accelerate the development of software widely.
In addition to human surveillance in the loop and limiting access to sensitive data, the majority of managers also access technology via trusted suppliers to cover themselves with risk, as indicated by 74% leaders in KPMG of Q3 investigation.
Companies often continue to adopt agentic AI with the help of existing partners. Athina Kanioura, Vice-President EVP and strategy and transformation manager in Pepsicounderlined the Importance of relations with technological suppliers with companies in interviews with CIO plunged earlier this year.
“Organizations strengthen their fundamental controls such as zero confidence, immutable forest exploitation, transparency of the thought chain and security protocols – not because technology requires it, but because the issues do it”, Mcgowan said in an email.
The Gartner analyst company considers that the addition of AI agents to operations leads to weaken upright cybersecurityReduce the time required to use authentication channels 50% over the next two years.
As companies strengthen governance, the potential benefits of agent systems are half.
The increase in risk attenuation and the increase in deployment of AI agents represent an intentional evolution, according to Mcgowan.
“When you deploy systems that can coordinate semi-autonomous or autonomous and adapt to fundamental principles in real time such as responsibility, transparency and security are no longer optional. Mcgowan said. “Organizations that get this problem understand that scaling up agency capacities and scope of scaling are not distinct initiatives.”
