The next evolution: agentic AI in modern tech teams

£78 billion. That‘s how much has been invested in the UK’s AI sector. But MIT suggests £74.1 billion of that will be wasted on pilots that never reach production. The research and our own experience demonstrate that the difference between the winners and losers is not rushing to build an impressive pilot – it‘s strategic execution.

 

Three critical failures emerge from working with enterprise clients: unclear business objectives that are divorced from AI capabilities, siloed implementations that don’t integrate with existing workflows, and a lack of governance frameworks for autonomous decision-making. Each is solvable, but only if addressed before deployment, not after.

 

As agentic AI transforms from experimental to operational, the companies that understand how to deploy, run it and optimise it successfully are opening an unbridgeable competitive gap. The question is quickly changing from “Will your organisation transform?” to “Will you lead the transformation in gaining tangible business value or be forced to react to competitors who moved first?”

 

Managing more complex workflows

 

Agentic AI is entering a new phase of capability, moving beyond simple automation to excel in complex, high-stakes workflows that demand nuanced decision-making and adaptability. The near future of AI will be defined by its ability to navigate intricate environments, co-ordinate across diverse systems and tackle challenges that were once considered too sophisticated for automation.

 

Take the legal field as an example. Agentic AI can analyse vast repositories of case law, identify relevant precedents and simulate potential outcomes based on different strategies. It can assist legal teams in crafting comprehensive case strategies, ensuring that every angle is considered while saving significant time and resources. This level of complexity requires AI to not only process data but also understand context, weigh competing priorities and adapt its recommendations as new information emerges.

 

This evolution in AI’s role is reshaping how organisations approach their most critical processes. By embracing agentic AI, businesses can unlock new levels of efficiency, innovation and resilience, positioning themselves to thrive in an increasingly complex and competitive world. But it’s important to note that as AI autonomy grows, human oversight will not be lost.

 

A new chapter for human-AI collaboration

 

As agentic AI moves into a more central role within IT teams a fundamental shift in required skillsets will emerge. The next generation of IT professionals will act less as technical specialists and more as strategic managers – overseeing a diverse stable of agents that operate autonomously, but also collaboratively, across multiple systems and business functions. The division between traditional specialisms will blur, with a greater premium placed on well-rounded, systems-oriented knowledge and the ability to orchestrate agents effectively.

 

Essential management capabilities will extend beyond people to include AI agents themselves. Future IT team members will become adept at setting rules for agent autonomy, defining how different agents interact and establishing policies for escalation and intervention. Crucially, even as agents become more capable, accountability for their outputs will remain firmly with their human supervisors.

 

Scrum cycles and stand-ups will persist, but with new dimensions focused on monitoring agent behaviour, evaluating outcomes and refining instructions. This evolution will increase organisational agility, allowing teams to deploy and redeploy AI resources fluidly as needs change and as new opportunities arise.

 

Dispelling the myth of replacement

 

It is a persistent misconception that advances in AI inevitably result in fewer opportunities for human workers. In practice, history has shown that disruption consistently brings new forms of employment and new responsibilities, and agentic AI will be no exception. Over the coming years, expect the emergence of roles focused on validating model logic and behaviour – requiring technical mastery and a deep understanding of business context.

 

Ethical implementation will become paramount. Testers will be called upon to develop robust new methodologies for detecting and eliminating bias in AI models and algorithms, ensuring applications operate equitably and responsibly for all users. These “guardians of ethical AI” will become a key pillar of every IT team, safeguarding against unintended discrimination and regulatory breaches.

 

Resilience advocacy will also take centre stage. As agentic AI systems become critical to business operations, the definition of resilience will shift: excellence will be judged not only on uptime, but on the stability, accuracy and ethical behaviour of AI outputs under stress. Testers and engineers will transition from verifying basic system operation to proactively identifying and mitigating instances where AI models risk producing biased, incorrect or unreliable results.

 

Democratising advanced automation

 

95 per cent of UK businesses are using AI or exploring it – but, as we‘ve seen, most of those initiatives will fail. This paradox defines the current moment: access to AI technology is democratised, but access to AI execution expertise is not. The barrier is strategic capability rather than technological availability.

 

As sizable language models age, they are likely to become freely or cheaply available, allowing smaller organisations to experiment with agentic AI using mature, proven technologies. These opportunities will enable leaner teams to deliver outsized improvements in customer support, supply chain management, fraud detection and more.

 

Empowered by early successes, such businesses will be well placed to justify further investment, incrementally automating more processes and leveraging AI as a catalyst for growth. However, realising these benefits is far from guaranteed, as the MIT research suggests. This underscores the critical importance of robust strategic planning and technical expertise throughout the adoption process. While democratisation will lower barriers to entry, organisations must approach AI with a clear vision, strong frameworks and rigorous oversight to ensure successful and sustainable innovation.

 

The agent economy is here. The technical capabilities exist. The business case is proven. What separates the 5 per cent who succeed from the 95 per cent who fail is the discipline to implement strategically, the frameworks to measure what matters and the expertise to turn pilots into production systems. That 18-month window to establish a competitive advantage is already closing.

---

Contact Ten10 to learn more about the impact AI and agentic automation can make in your business