A study by Gartner shows that only 15% of IT application leaders are considering, piloting, or deploying fully autonomous AI agents within their organization.
This study, which attracted insights from 360 IT application leaders in North America, Europe and Asia/Pacific, found that while the majority were involved with AI agents to some extent, fully autonomous implementations remain limited. This study focused on the deployment, trust, and perceived impact of generated and agent AI across enterprise applications.
Adoption and barriers
Note that while 75% of respondents have pilots, deployments, or already deployed AI agents in some way, concerns about governance, solution maturity, and agent sprawls have slowed down the move towards full agent AI. These agents are defined as goal-driven AI tools that do not require human surveillance, and often face pushback due to issues of security and organizational preparation.
Trust remains an important barrier. Only 19% of respondents reported high or complete confidence in the vendor's ability to provide adequate hallucination protection. Hallucination protection refers to a mechanism that prevents an AI system from generating false or misleading output. Furthermore, 74% believe that AI agents are a new attack vector, and only 13% strongly agreed that organizations have an effective governance framework in place to manage new risks associated with agent AI.
Max Goss, Gartner's senior director analyst, explained the situation.
“The hype around agent AI continues to grow, and vendors position AI agents as the next step in AI evolution in addressing the lack of more traditional Genai assistants. 75% of survey respondents say they are piloting, or have already deployed them in some form of AI agents.
Impact assessment
The findings show tempered expectations of transformative change driven by AI agents. Only 26% of respondents felt that agents had a transformative impact on productivity. A majority of 53% said the impact was significant but may not be transformative, while 20% thought the profits were small.
Organization-wide alignment appears to play a role in perception of the impact of AI. Only 14% strongly agreed that there is clarity between it, business users, and leadership about the issues that AI should solve. Those with greater adjustment were 1.6 times more likely to view AI agents as transformative, and more than three times more likely to extract significant value from the generated AI tools.
Goth emphasized the importance of this alignment.
“The integrity between business and executive leadership over the problems that AI can solve and how it can measure its value is important to successful deployment of AI, but we can see that many organizations don't.”
Deployment focus
Many organizations may not concentrate their AI deployment in the most effective areas. People with less business alignment are almost twice as likely to see office productivity, like the top domain of AI agents, but organizations with clearer goals tend to prioritize vertical use cases such as customer service, ERP, and sales.
As Goth said,
“Office productivity and digital workplaces are the defaults for organizations that don't have a strong grasp of what they're doing with agents, but they're not necessarily the areas offered to organizations to the most valuable organizations. Of all survey respondents, they were top of the list, with 64% being ranked as AI agents being the most affected and customer service (55%) the most affected (55%).
Exchange of applications and labor
According to most leaders, the exchange of applications and workers by AI agents within the next two to four years remains unlikely. Only 12% of agents strongly agreed to replace applications, while only 7% strongly believed that they would replace workers within that time frame. However, considering those who agreed slightly, the proportion rose. 34% believe that applications will be exchanged, while 29% expect workers will replace AI agents in the coming years.
Goth commented,
“This remains important for technology that has been commonly available over the past 12 months. It points to both the hype and fear that exists around AI, particularly agent AI.”
Recommendations for organizations
Gartner analysts advise organizations to focus on three key areas for agent AI deployment.
- AI Agent Governance: Build a platform-independent governance framework to reduce sprawl and set clear policies for safe use of AI agents across a variety of tools and domains.
- Impact Domain Target AI Agent: By achieving alignment between IT and business teams about specific issues that AI should address, you increase the likelihood of achieving value from these technologies. If risk and returns are uncertain in office productivity scenarios, organizations may find significant benefits in customer service or analytics applications.
- Adopt a multi-vendor strategy: Given the current immaturity of technology, organizations should not rely on a single vendor of AI agent solutions, instead assessing multiple options across their ERP, CRM, and digital workplace portfolios.
Gartner's report concludes that Agent AI adoption remains in the early stages, focusing on developing governance structures and identifying high-value use cases across the enterprise.
