The rise of AI agents and their impact on work
While the hype and pace of change that artificial intelligence has brought might seem really fast to many, this breakneck evolution may have been just the initial proof of concept pointing to where AI will have its real and lasting impact.
That’s because AI is moving past traditional use cases focused on chatbots and prompting towards environments dominated by AI agents – automated, proactive systems capable of independent decision-making. From September to October 2025, Aberdeen Strategy & Research conducted a 15-minute online survey among 281 IT decision makers who are personally involved with tech purchase decisions related to AI and AI agents for their organization. The study examined how IT professionals across North America, EMEA, APAC, and India are using AI and agents today and the outcomes they’re achieving. Using this data, Aberdeen uncovered how businesses of all sizes are navigating the transition to AI agents, highlighting the drivers, adoption patterns, and measurable successes of agentic systems.
Using this data, Aberdeen uncovered how businesses of all sizes are navigating the transition to AI agents, highlighting the drivers, adoption patterns, and measurable successes of agentic systems. Also, while many studies have shown low levels of success with some AI implementations, when Aberdeen looks at businesses who have taken practical approaches to AI with clear focus and end-goals, our research shows much higher levels of AI success.
Unlike older scripts or “daemons” that strictly follow pre-specified, “if-then” rules, AI agents are characterized by their autonomy. They respond to environmental changes, different data triggers, and new information by making their own decisions to achieve a goal.
Key characteristics of these systems include:
- Autonomy: They can operate completely without human intervention.
- Collaboration: They can be designed to communicate and work with other AI agents in an agentic system.
- Real-world Application: A classic example of this technology in action is the autonomous vehicle, which must constantly process environmental data and make split-second decisions without a human prompt.
Business leaders are not pursuing AI agents merely for the sake of innovation; they’re looking for specific operational improvements. According to Aberdeen research, automating IT processes is the primary driver, cited by 67% of respondents. This is followed by the desire to improve customer support (58%).
Beyond these top goals, organizations are leveraging agents for:
- Predictive Analytics and Forecasting: 46% of firms use agents to better anticipate events, disruptions and trends.
- Cybersecurity: 46% see agents as a vital tool for improving security postures through proactive monitoring and speedier responses.
- Sales and Supply Chain: Respectively, 40% and 33% of firms are using agentic systems to boost sales effectiveness and automate supply chain logistics.
The data also shows that the larger the company, the more integrated AI agents have become. Approximately half of all midsize and large companies report they now use AI agents throughout their organizations.
Small companies are also finding success, though their implementation may be more targeted. Interestingly, despite often having fewer resources, 77% of small companies report that their AI solutions are meeting ROI expectations. Furthermore, 72% of small companies rate the performance of their AI solutions highly, and 72% are satisfied with system uptime.
The effectiveness of AI agents is heavily dependent on the underlying IT infrastructure. Organizations that are currently using AI agents have significantly higher implementation rates of specific supporting technologies.
The survey identifies key infrastructure components that support these initiatives:
- Deeper Adoption: Accuracy of AI capabilities is directly linked to the depth of technology adoption. Firms with high adoption see better results than those with low or medium adoption.
- Agentic Business Support: Businesses successfully running agents often invest more heavily in robust data pipelines and specialized hardware to handle the more demanding nature of agentic workloads.
A critical metric for any new technology is engagement. Our research shows that large companies currently lead the way in both employee and customer engagement with AI.
- Employee Engagement: 84% of employees in large companies interact with AI everyday or regularly, compared to 77% in small companies.
- Customer Engagement: The gap is even wider in customer interactions. 91% of customers at large and mid-size firms engage with AI frequently or occasionally, while only 80% do so at smaller firms.
Accuracy is another area where deepened adoption pays off. Among firms with high technology adoption, the accuracy of their AI capabilities is rated as excellent or good by 93% to 96% of respondents. This suggests that as firms move past the pilot phase and into full deployment, the reliability of the agents increases substantially.
As we move further into 2026, the focus for many organizations will likely shift from whether to use AI agents to how to better integrate them into complex workflows and improve the underlying infrastructure to maximize accuracy and engagement.
