60% of AI Agents in IT Departments: Daily Tasks Revealed

The Rise of AI Agents in Enterprises

AI agents are becoming the next big thing in the world of technology, and it seems like everyone wants a piece of the action. But what exactly are these agents up to within businesses? In many cases, they're not just working on their own tasks but are also helping to create more agents. They're particularly useful in IT departments, where they manage system performance, including the very infrastructure that supports other AI agents. However, their roles can vary widely across different industries.

A recent survey by Cloudera, involving 1,484 IT leaders, revealed that a staggering 96% of organizations are gearing up to expand their AI usage over the next year. That's an impressive figure, considering that typically at least 10% of respondents in any survey are outliers. Moreover, 57% of these organizations have already started implementing AI agents within the last two years. Yet, the excitement is tempered by concerns over data privacy, integration challenges, and data quality issues, which could potentially derail these plans.

Current Applications and Deployment Strategies

Most AI agents currently in use are embedded within IT operations, with the majority focused on performance optimization (66%), security monitoring (63%), and assisting in development (62%). So, where do these agents come from? A significant 66% of respondents are building these agents using enterprise AI infrastructure platforms, while 60% are tapping into agentic capabilities already embedded in their core applications. According to the survey's authors, this approach highlights a preference for deployments that are scalable, secure, and closely tied to data sources.

Beyond IT optimization, AI agents are making their mark in customer-facing operations, primarily in customer support (78%), process automation (71%), and predictive analytics (57%). When it comes to building these agents, the technologies of choice include enterprise AI infrastructure platforms (66%), agent capabilities within applications (60%), and dedicated AI agent platforms and frameworks (60%).

Challenges and Concerns

Of course, AI agents are not without their issues. Common challenges include data privacy concerns (53%), difficulties integrating with existing systems (40%), and high implementation costs (39%). Over a third (37%) of respondents found integrating AI agents into their current systems and workflows to be "very" or "extremely" challenging. As the survey authors noted, deploying AI agents is far from a simple plug-and-play process, echoing the persistent challenges seen with previous technological advancements.

Tech leaders are eager for improvements in AI agents, particularly in areas like data privacy and security features (65%), faster training and customization options (54%), better natural language processing (51%), and improved contextual understanding (50%).

Industry-Specific Use Cases

The use cases for AI agents vary significantly across industries:

• In finance and insurance, the leading applications include fraud detection (56%), risk assessment (44%), and investment advisory services (38%).
• Manufacturing sees AI agents being used for process automation (49%), supply chain optimization (48%), and quality control (47%).
• In healthcare, AI agents are primarily used for appointment scheduling (51%), diagnostic assistance (50%), and processing medical records (47%).
• Telecommunications leverage AI agents for customer support bots (49%), enhancing customer experience (44%), and security monitoring (49%).

Recommendations for Implementing AI Agents

The Cloudera survey provides several recommendations for businesses looking to implement AI agents effectively:

• Strengthen Data Foundations: Ensure you have a modern data architecture and unified platforms capable of handling the diverse and voluminous data required by AI agents.
• Focus on High-Impact Projects: Start with initiatives that promise immediate returns on investment, such as customer support and process automation, which directly address critical business needs.
• Establish Clear Accountability: Define who is responsible for the agent's performance—whether it's the developer, the business owner, or the operations team.
• Develop Governance and Ethics Frameworks: Implement systems to audit bias, ensure transparency in decision-making, and regularly assess agent behavior against company policies and user expectations.
• Upskill Teams: Move beyond basic training to develop hybrid skill sets that allow employees to build, integrate, and understand AI agents, fostering a culture of continuous learning and collaboration between humans and AI.

The enthusiasm reflected in the survey responses suggests that AI agents are poised to be the next major wave in AI technology, offering targeted solutions rather than the complex, overarching systems that many feared. It will be fascinating to see if the planned 96% adoption rate becomes a reality.