UK firms embed AI in workflows but keep human control

Latest UK research from SAP Engagement Cloud suggests enterprise AI is becoming embedded in business workflows, while human oversight remains firmly in place.

The findings are based on a survey of more than 750 UK enterprise IT decision-makers. Among them, 79% said AI-driven assistants had increased productivity without reducing human control, while 35% strongly agreed that AI is now embedded in business workflows rather than operating as a standalone tool.

The results point to a shift in how large organisations use AI in customer engagement and wider operations. Rather than limiting it to chatbots or separate tools, businesses are increasingly integrating AI into routine processes that shape customer interactions and internal work.

Governance remains central to that shift. In the survey, 78% of respondents said their organisation had clear AI guardrails covering data lineage, personally identifiable information handling, and human approval points.

The same share, 78%, said their organisations were making significant investments in AI-powered customer engagement. This suggests spending is continuing even as companies face pressure to show AI systems can operate within established controls.

Sara Richter, Chief Marketing Officer at SAP Engagement Cloud, said the issue for many businesses is no longer whether AI is useful, but where it should be allowed to operate autonomously.

"Most businesses already know that AI can help them. The question now is where it should be allowed to act, and where people still need to stay in control," Richter said.

She added: "AI agents can remove a lot of the slow, manual work that obstructs good customer engagement. But they are only successful when the data is connected, the rules are clear, and teams trust what the AI is doing."

Governance focus

The emphasis on controls reflects broader market concerns about accountability for AI-driven decisions. As companies deploy AI in customer-facing and operational workflows, they must also manage how data is sourced, how outputs are approved, and where responsibility sits when automated systems influence outcomes.

Professor Mark Ritson described AI as a tool that still requires human direction.

"AI is the racehorse, not the jockey. It's remarkably effective at covering ground quickly, but it still needs someone holding the reins. The best companies are using AI to accelerate execution while keeping strategic decisions, judgement, and accountability firmly in human hands," Ritson said.

The survey also suggests connected data remains a practical barrier to broader deployment. SAP linked successful AI adoption to the quality of connections between customer information and operational systems such as inventory, fulfilment, service, loyalty, and finance.

This matters because disconnected systems can lead to inconsistent customer experiences. If AI recommends an action a business cannot support operationally, the technology may expose weak coordination rather than improve service.

Data challenge

Richter said poor data and fragmented systems can undermine the value of AI.

"AI agents are incredibly effective. But if your data is messy or your systems don't talk to each other, layering AI on top will only make those problems more visible," she said.

"The brands that get this right will be those that start with strong foundations: connected data, clear permissions, and workflows that improve as they learn. That's how you move faster with AI without losing control," Richter added.

SAP also pointed to early adoption by brands seeking to combine customer insight with operational context. One example was Jack Wolfskin, which is exploring how AI can bring together customer data and surface information that helps teams tailor interactions more closely to behaviour and channel preference.

Michael Walter, Senior Direct Marketing Manager at Jack Wolfskin, said the company saw scope to use AI and data insight to tailor journeys across touchpoints.

"One of the biggest opportunities we see is combining data insight and AI to personalise journeys in real time across all touchpoints," Walter said.

He added: "We can respect channel preferences, trigger communication based on behaviour, and use AI-enhanced recommendations for next-best actions, offers, and context-aware engagement. For our customers, this means communication that feels relevant and personal; for us, it means higher engagement, stronger loyalty, and more efficient use of every channel."

The research covered decision-makers in IT, technology, marketing, service, and revenue roles at UK businesses with more than 500 employees and annual turnover above USD $250 million.