Artificial intelligence is accelerating across every part of the digital construction landscape, but confidence in its use is not keeping pace. New research shows that public concern about AI is rising sharply, with fears centred not on job loss, but on how AI is governed and controlled. These findings mirror the challenges many enterprises face as they adopt AI without the structures, data foundations, and oversight needed to manage the technology safely.
A joint study by Cybernews and nexos.ai charted AI-related concerns in the US from January to October 2025, analysing trends across governance, privacy, employment, misinformation, and bias. The result was clear: the strongest public anxiety relates to control and regulation, followed closely by data and privacy. Concerns about job displacement ranked lowest, even during a year of widespread layoffs across the tech sector. For organisations in the UK and beyond, these findings highlight a widening trust gap that sits at the heart of enterprise AI adoption.
Žilvinas Girėnas, from nexos.ai, describes the tension facing business leaders. The fear is not AI itself, but the loss of visibility when unapproved tools begin appearing across teams. The rise of ‘shadow AI’, where employees use external AI tools without governance, creates uncertainty about where company data is going and how decisions are being made. Without visibility and oversight, organisations cannot manage compliance, security, or operational risk.
Where AI anxiety begins
The study identifies several root causes of AI anxiety that apply equally across the public and enterprise environments. One key factor is the opacity of advanced AI models. Many operate as ‘black boxes’, producing outputs without clear reasoning. This lack of transparency heightens concern about losing control of systems that are increasingly embedded in critical decision-making.
Data collection is another source of unease. AI systems are trained on extensive datasets, often incorporating information gathered without explicit consent from digital platforms, online behaviour, and connected devices. This increases fears around identity theft, personal data misuse, and the erosion of privacy.
Misinformation also contributes to a broader distrust of AI. Highly realistic synthetic content blurs distinctions between real and fabricated information. When AI models learn from biased datasets, they risk amplifying long-standing inequities, especially in areas such as recruitment, finance, and public services. Although job displacement ranked lowest in the study’s anxiety categories, the psychological impact remains significant, with concerns focused on purpose, identity, and long-term career stability.
Girėnas argues that organisations face a similar dynamic: when teams do not fully understand how AI works, confidence drops and adoption slows. He suggests that trust must be built through visibility and governance across the entire AI ecosystem.
The tangible risks of losing control
The enterprise sector is already experiencing the consequences of uncontrolled or ungoverned AI use. According to a recent McKinsey analysis, organisations deploying AI are encountering several recurring risks.
- Inaccuracy and reputational harm: The most common negative impact reported is inaccurate output. Unvetted AI tools can introduce incorrect or fabricated information into reports, client communications, or automated systems, damaging trust and credibility.
- Cybersecurity vulnerabilities: Over half of organisations using AI are actively working to mitigate cybersecurity risks. The concern is that AI tools may leak sensitive data or be exploited to introduce malicious code into operational systems.
- Intellectual property exposure: Many organisations fear that proprietary code, architectural plans, or strategic data entered into public AI tools may be absorbed and reused. Companies with advanced AI adoption are more likely to report IP infringement, reflecting the risks faced when data governance is unclear.
- Compliance failures: Without centralised AI governance frameworks, organisations struggle to meet regulatory requirements such as GDPR and the emerging EU AI Act. This increases the risk of legal and financial penalties.
UK businesses face a readiness challenge
These risks are amplified in the UK, where rapid AI adoption is colliding with outdated technology estates. Analysis from Propel Tech shows that many organisations face structural barriers to AI integration, including legacy software, siloed data, and fragmented platforms. While these issues may have been manageable in conventional environments, AI’s demand for consistent, real-time data exposes them immediately.
Recent S&P Global Market Intelligence findings support this trend: AI project failure rates have doubled, and 42% of organisations have abandoned most AI initiatives due to underestimated complexity. This gap between ambition and readiness is contributing to enterprise AI anxiety, from boardrooms to frontline staff.
The cultural impact is equally significant. Research from Propel Tech and the TUC reveals widespread concern among UK workers about AI’s impact on jobs, particularly among those aged 25–34. At the same time, employees with AI skills are experiencing substantial wage growth, increasing the pressure on organisations to adopt AI even when systems are not prepared.
In many cases, AI failures stem not from the technology itself but from deploying it on ageing, disconnected systems. Without robust data pipelines, standardised documentation, or integrated platforms, AI cannot operate reliably. This creates uncertainty for leaders, employees, and technical teams alike.
Building readiness and closing the trust gap
Industry analysts agree that organisations must prioritise readiness before accelerating AI adoption. This includes governance frameworks, system audits, data-quality assessments, and clear operational processes. Girėnas recommends four core actions for leaders:
- Centralise AI governance to ensure consistent rules across the organisation.
- Introduce human-in-the-loop processes to maintain oversight of high-impact decisions.
- Elevate AI governance to the C-suite, embedding accountability at the highest level.
- Shift from restriction to visibility, focusing on understanding which tools teams are using, rather than simply banning them.
Propel Tech echoes these recommendations, arguing that readiness assessments and software modernisation are essential steps for reducing risk and enabling long-term value from automation.
Andrew Brown, Director at Propel Tech, summarises the challenge: AI is highlighting longstanding weaknesses in enterprise systems. The anxiety is not driven by the technology itself, but by the readiness of organisations to adopt it safely and effectively.
As AI becomes integral to digital construction and wider business operations, the trust gap between ambition and preparedness is widening. Organisations want the benefits of automation, insight, and operational efficiency, but many lack the governance, data quality, and system integration required to adopt AI responsibly. Addressing these foundations is now essential. The companies that invest in modernising systems, establishing visibility, and building enterprise AI governance frameworks will be best positioned to innovate with confidence and reduce the anxiety surrounding AI adoption.
Read next: Advisory council to accelerate low-carbon cement innovation – Build in Digital
Are you a building professional? Sign up for a FREE MEMBERSHIP to upload news stories, post job vacancies, and connect with colleagues on our secure social feed.
