Numerous surveys paint a stark picture: businesses worldwide are failing to translate their artificial intelligence (AI) investments into measurable returns. A big part of the problem, argues Bernhard Schaffrik, principal analyst at Forrester, is that AI providers are systematically ignoring the human impact of their technology. Speaking at the CamundaCon 2026 conference in Amsterdam, Schaffrik highlighted that AI vendors often focus narrowly on technical capabilities while neglecting how their tools affect employees, workflows, and corporate culture. "They completely ignored the human factor, and also the enterprise factor," he told Computer Weekly.

The fear among employees that AI will replace their jobs is palpable. In its 2025 paper, Building a pro-worker AI innovation strategy, the Trade Union Congress recommended that employers ensure meaningful worker participation at every stage of AI deployment—from strategy development and problem definition through tender, application design, and deployment. The report argues that involving workers not only addresses ethical concerns but also drives the effectiveness of technology. Yet, many organisations skip this step, leading to resistance and underutilisation.

While change management has been a staple of corporate transformations for decades, Schaffrik points out a critical difference with AI—especially generative AI. Because AI tools are so accessible and seemingly easy to deploy, the conversation often starts at the boardroom level. CEOs immediately grasp the potential for cost savings, efficiency gains, and competitive advantage, and they push implementation aggressively into their organisations. "CEOs assume that people down the line—HR teams, transformation leaders—will handle change management," Schaffrik said. However, he warned that AI directly threatens job roles, causing exponential increases in employee fear and confusion. Those tasked with implementing AI are often themselves impacted by it, creating a cycle of resistance and uncertainty.

Schaffrik also highlighted the rigidity of enterprise technology frameworks and business processes as a major obstacle. Businesses cannot simply replace a mission-critical system like payroll without careful risk assessment. "Businesses don't want to break the payroll process," he said, explaining that AI providers are often surprised when their technology fails in real-world deployments. The surprise stems from a combination of human psychology, organisational inertia, and regulatory constraints. These issues are not new, but AI amplifies them because of the speed and breadth of change it demands.

To navigate these challenges, Schaffrik advises CEOs to clearly define what AI should accomplish. He advocates automating as much repetitive work as possible using any enterprise-grade technology—workflow engines, robotic process automation, document processing, and AI agents. At the same time, leaders should target processes that are not highly repeatable but are prone to human error. For example, a legal professional comparing multiple versions of a lengthy contract may make mistakes; AI, despite its tendency to hallucinate, can perform such comparisons more reliably when properly supervised.

This is where the human-in-the-loop concept becomes essential. Schaffrik notes that employees who serve as AI output checkers must excel at the very tasks AI is taking over. A legal expert reviewing AI-generated contract comparisons must be highly skilled in contract analysis. As AI evolves, upskilling will be crucial—not only to maintain quality but also to justify higher salaries for these oversight roles. Organisations that invest in training their workforce to work alongside AI are more likely to see positive outcomes.

Historical context supports this view. The industrial revolution, the rise of the internet, and the advent of cloud computing all required significant workforce adaptation. AI represents a similar inflection point, but with a twist: it affects cognitive work as much as manual labour. Early adopters of AI in sectors like finance, healthcare, and manufacturing have learned that successful integration depends as much on culture and communication as on technology. Those that fail to address employee fears often face low adoption rates, shadow IT (where workers use unauthorised tools), and outright sabotage.

Schaffrik’s insights align with broader research. A 2024 McKinsey report found that companies with strong change management practices were three times more likely to report successful AI deployments. Similarly, Gartner has noted that through 2026, half of AI initiatives will fail due to lack of attention to human factors. The message is clear: technical excellence alone is insufficient.

For CEOs, the path forward involves several concrete steps. First, establish transparent communication about AI’s purpose and limitations. Second, involve employees early and often in design and rollout. Third, create clear career paths for workers whose roles evolve—such as from data entry to AI oversight. Fourth, invest in robust feedback loops to continuously improve AI systems based on real-world usage. Finally, partner with AI vendors that prioritise enterprise integration and human-centred design.

The AI revolution is inevitable, but its success hinges on how well organisations manage the human side of disruption. As Schaffrik’s analysis shows, businesses that ignore the human factor will continue to see their AI investments fail to deliver. By combining strategic automation with empathetic change management, leaders can turn AI from a source of fear into a driver of productivity and innovation.

Source: ComputerWeekly.com News