**London**: Industrial enterprises are poised for a surge in agentic AI adoption from under 1% in 2024 to 33% by 2028 but face significant barriers such as change management, data readiness, workforce resistance and unclear metrics delaying full-scale deployment.
In the rapidly evolving landscape of industrial enterprise technology, agentic artificial intelligence (AI) is poised for significant growth. According to Gartner, the incorporation of agentic AI in enterprise software applications is expected to surge from less than 1% in 2024 to 33% by 2028. Industrial enterprises are increasingly interested in leveraging this transformative technology to enhance real-world business applications. However, many firms encounter a challenging phase often referred to as “Proof-of-Concept (PoC) Purgatory,” where AI initiatives fail to progress beyond pilot testing to full-scale deployment.
Bob De Caux, Chief AI Officer at IFS, highlights that this transition is complex and unlike previous technology adoptions. Several barriers contribute to the difficulty in scaling agentic AI, which refers to autonomous AI agents operating broadly within business processes.
One major obstacle is change management. Industrial companies tend to be cautious about adopting technologies that significantly alter their operations due to fears of losing control over critical processes and uncertainty regarding the impact on established workflows. This hesitation often results in delayed or abandoned projects.
Another challenge is the absence of clear success metrics. Enterprises struggle to define how agentic AI’s impact on productivity, cost reduction, and operational efficiency should be measured. Without these metrics, it becomes difficult to make informed decisions about the value and progress of AI tools.
Identifying suitable use cases also presents a significant barrier. Industrial firms must possess deep domain expertise and insight into their workflows to determine which complex processes can be effectively managed by agentic AI. Frequent attempts to test simplistic or non-representative scenarios result in stagnation within the testing phase and limit the technology’s potential.
Data framework readiness is equally critical. Despite 86% of organisations acknowledging data readiness as essential for AI success, only 23% have established the necessary infrastructure. Legacy technology systems, fragmented data, and outdated processes create complications that are magnified with the introduction of agentic AI, which generates large volumes of real-time data requiring robust data processing capabilities.
Workforce resistance also plays a pivotal role. Early feedback from deployments such as chatbots indicates that employees often push back against automation that removes certain manual tasks. Adapting to the shift where humans oversee AI-driven outcomes rather than execute every step demands workforce engagement and adjustment.
To overcome these challenges and move towards scaled AI deployment, enterprises should consider several strategic steps:
-
Define clear business objectives linked to specific agent roles and establish relevant key performance indicators (KPIs). For instance, background monitoring agents might focus on uptime improvements, while end-to-end process automation agents target productivity gains.
-
Ensure data and infrastructure readiness by upgrading systems, mapping processes clearly, and implementing governance frameworks to maintain security, compliance, and operational reliability.
-
Adopt a phased rollout approach that begins with targeted, impactful agents. Iterative refinement based on feedback enables adaptation to real-world conditions and smoother scaling.
-
Engage employees through comprehensive training and upskilling to facilitate workforce alignment. Early involvement of cross-functional teams—including operators, IT, and business leaders—helps build collaboration and ownership.
-
Continuously monitor and evaluate agent performance against set KPIs, adjusting autonomy levels as appropriate. Establishing enterprise-wide frameworks supports streamlined future deployments and enhances overall confidence in scaling.
The Tech Radar report emphasises that while many organisations have begun deploying standard AI agents, achieving fully autonomous agentic AI at scale remains a substantial endeavour. Strategic investments and methodical deployment models are essential for moving beyond PoC phases and realising tangible operational benefits before Gartner’s predicted widespread adoption by 2028.
This analysis was provided as part of TechRadarPro’s Expert Insights channel, showcasing insights from leading voices in technology today.
Source: Noah Wire Services
Subscribe to Industry Updates
