While each organisation is at a different point in its journey, there was shared recognition that AI is becoming central to operational efficiency, innovation and future competitiveness.

A prevailing theme throughout the conversation was the extraordinary pace of change in AI capabilities and tooling. For many, traditional planning and investment cycles are already outpaced. New models, use cases and tooling are surfacing monthly, if not weekly. This reality presents a stark choice:

Organisations that aren’t actively experimenting with AI now risk falling behind, both in productivity and market relevance. Balancing early adoption underpinned by governance frameworks, which prepares teams for the very near future, was frequently cited as critical to future readiness.

Participants also agreed that deploying cutting-edge models alone does not guarantee a successful AI transformation. What needs to take precedence is embedding the right capabilities into the organisation with a responsible approach that is secure by design and clear on business outcomes. This set the stage for a wide-ranging discussion, shaped around a series of key questions and themes that reflect where organisations are and where they’re headed.

‍

‍

1.    Are AI co-pilots a safe first step into automation?

‍

Yes - the initial focus should be on augmentation, not replacement. Some attendees are beginning to explore the use of AI-powered co-pilots to support complex, knowledge-heavy operational tasks. These early experiments are focused on augmenting human expertise, helping staff navigate complex systems and capturing institutional knowledge that is often undocumented.

‍

These tentative steps highlight a growing awareness that AI can support, rather than replace, human judgment in highly variable environments. “Co-pilots” provide a safe, incremental path toward automation while unlocking immediate value through improved consistency, training and task execution.

‍

‍

2.    How do we ensure AI is used responsibly across the organisation

‍

Several attendees at the roundtable realised the potential of establishing AI Councils or Committees to provide strategic oversight. These are typically cross-functional by design, bringing together voices from legal, HR, cybersecurity and operations to guide development, deployment and use.

This approach reflects a growing understanding that trust is a critical enabler of AI adoption for the organisation and in the eyes of regulators, customers and the public. In regulated industries, especially, proactive governance is essential to avoid reputational, ethical or compliance risks. Moreover, involving a broad set of stakeholders early not only strengthens oversight but also helps reduce resistance to change, improving buy-in across teams and functions.

‍

‍

3.    Why are AI pilots failing to deliver on expectations?

‍

There was widespread agreement that a lack of AI literacy and training is a bottleneck. Early pilots with Microsoft 365 Copilot often underperformed due to insufficient user enablement. Companies are now looking to roll out tiered, role-based training models as they recognise that culture change is as important as the technology and often the biggest barrier to transformation. To scale AI effectively, staff must understand its value, feel confident using it and see it as a tool that helps them rather than replaces them.

‍

4.    Should we prioritise innovation or financial discipline in AI investment?

‍

Both - Innovation and fiscal responsibility are not mutually exclusive, but must be intentionally balanced. Several attendees emphasised that for AI initiatives to gain lasting traction, they must be tied to clearly defined business outcomes, not pursued in isolation. There was strong agreement that “AI for AI’s sake” is a common pitfall, particularly when pilots are not linked to strategic KPIs or value creation. To avoid this, organisations are adopting structured governance and prioritisation frameworks that help identify where experimentation can deliver meaningful impact, without losing sight of cost control and resource efficiency.

Ultimately, sustainable transformation requires that AI investments stand up to the same scrutiny as any other business initiative: value-driven, outcome-focused and strategically aligned.

‍

5.    Vendor Lock-in vs. Open Architecture

‍

Concerns were raised around heavy dependency on a 'single vendor enterprise' ecosystem, with several attendees highlighting difficulties in resisting “default” big-player pathways. While each organisation must weigh its own priorities, there was a shared recognition in the room that over-reliance on a single vendor risks limiting long-term flexibility. As the AI market continues to evolve at pace, maintaining a model-agnostic, modular approach will be key to preserving choice, fostering innovation and avoiding technical or commercial lock-in that could constrain future strategy.

‍

Ntegra’s Perspective

At Ntegra, we feel particularly strongly about the need to continue to experiment with novel technology and develop strategies based on experiential learning. We believe that this should include a focus on:

• Partnering with innovators and research-led vendors to stay abreast of emerging capabilities and identify potential early-mover advantages.

• Framing agentic AI pilots around augmentation, not automation, with a focus on enhancing human decision-making, surfacing latent knowledge, or managing complex tasks across systems.

• Evaluating outputs rigorously with clear metrics and human oversight to ensure trustworthiness and relevance.

Importantly, we place value on working across sectors to develop a peer group to capture and share insights across teams to accelerate organisational learning and avoid duplicated effort.

The conversations throughout the Roundtable event highlighted an encouraging mix of ambition, pragmatism and shared learning. While no two organisations are at exactly the same point, the collective focus on human-centric, ethically governed AI reflects a mature and grounded approach. By continuing to build skills, platforms and governance in parallel, these organisations are well-positioned to unlock transformative value from AI in the years ahead.

We know no one has all the answers when it comes to navigating the deep and murky waters of AI but what’s clear is that sharing ideas, challenges and practical experiences makes a real difference so if you'd like to be part of future conversations like this, we'd love to hear from you.

‍

Connect with Ben Parish, Head of Innovation

‍