‘Without strong data foundations and clear governance, confidence in AI can outpace trustworthiness’
Human resources professionals are facing mounting people and governance risks as organisations lean more heavily on AI while controls and workforce protections struggle to keep pace, according to two studies.
A public‑sector report by analytics firm SAS and research company IDC warns that government agencies are moving rapidly from AI pilots to operational use without matching safeguards.
According to Data and AI Impact Report: The Trust Imperative, only 6% of government agencies are in an “ideal” state, combining high internal confidence in AI with systems that are demonstrably trustworthy.
The report — with regional analyses of North America, Europe, Latin America, META (Middle East, Turkey and Africa) and Asia-Pacific — finds that 38% of government organisations are “overrelying on AI while underutilising trustworthy AI safeguards,” particularly around generative AI.
SAS and IDC describe this gap as a “trust dilemma”: organisations either underuse reliable AI or place too much confidence in tools that have not been fully validated. The authors say public‑sector AI maturity is outpacing investment in trustworthy AI, increasing the risk of biased outcomes, security breaches and costly operational failures.
“For the public sector to rely on AI, it must deliver clear value while protecting the well‑being of citizens,” said Grant Brooks, Senior Vice President of Public Sector at SAS. “Realising this value for citizens and communities requires aligning AI ambition and readiness. The report findings suggest we have work to do to achieve that.”
Chris Marshall, Vice President for Data, Analytics, AI, Sustainability and Industry Research at IDC, added: “Without strong data foundations and clear governance, confidence in AI can outpace trustworthiness, increasing risk for citizens and agencies alike.”
Mandated use of AI tools at work is emerging as a potential driver of resignations, as employees push back against AI systems, according to a previous report.
Errors, bias and workforce fallout
The overreliance issue extends beyond government. A separate report from Gallagher shows employers are worried about both technical and people‑related fallout if AI systems fail.
AI errors, misinformation and hallucinations top the list of perceived threats from AI, followed by legal and reputational risk from misuse and privacy violations, according to the company’s 2026 AI Adoption and Risk Survey of 1,250 organisations in Canada, the US, the UK and Australia.
Overreliance and reduced human judgement also feature, alongside job insecurity, potential industrial action, declining employee engagement and change fatigue.
John Farley, Managing Director of Cyber at Gallagher, warns that AI‑driven exposures are qualitatively different from traditional IT incidents. “AI liabilities aren't simple data breaches; they're a black box of algorithmic risk where traditional breach response approaches fall short,” he said, arguing that organisations need multidisciplinary oversight that blends governance, bias controls and data integration.
Yet employers are seeing significant upside. Gallagher finds 86% of businesses believe AI has improved employee productivity and 82% report a positive impact on revenue. Nearly two‑thirds of organisations measuring return on investment expect it will take about 28 months for AI benefits to outweigh costs. That mix of gains and emerging risks is creating a pivotal moment for HR and senior leadership.
Nearly half (46%) of women who use AI in their jobs are concerned about facing negative consequences from doing so, according to a previous report.
Training and governance
To capture value and contain risk, employers are ramping up skills investment. Gallagher’s survey shows nearly two‑thirds of businesses (62%) delivered AI training to employees in the past year, while 47% now offer specific training on AI tools, up seven percentage points from 2024. More than half (55%) have hired for AI‑focused roles and 40% have created jobs where AI is core to the remit.
“For many global companies, AI is no longer in the test phase. It's in the workplace, shaping strategy and powering productivity,” says Ben Warren, Managing Director of People, Data, AI and Innovation at Gallagher. “Training programs are on the rise, equipping employees for a future where human ingenuity and AI agents will work hand in hand.”
Despite the value of AI, Gallagher’s research indicates employers still see human capability as central. The most common reason for job‑protection strategies is to retain and promote creativity; other key reasons include preserving the human touch in client interactions and ensuring people remain available for complex problem‑solving and high‑stakes decisions.
However, governance is not advancing at the same pace. Gallagher finds 56% of organisations have communicated their AI strategy to employees, but 43% have yet to introduce formal AI risk‑management frameworks and only 44% have carried out AI impact assessments. Fewer than half have AI‑specific incident response plans.
The SAS/IDC report highlights similar weaknesses. Across all regions, government respondents cite a lack of centralised or optimised data foundations as the top challenge to AI implementation, followed closely by weak data governance. Governments expect strong AI investment growth, but lag banking and insurance on SAS’s Trustworthy AI Index and on spending directed specifically at trustworthy AI.
Both reports conclude that governance and accountability must quickly catch up with adoption. Lenin Lopez, Senior Vice President for Management Liability at Gallagher, says boards need to ask, “To what extent do we want and/or need to put controls in place?” and ensure leaders have enough AI knowledge to evaluate initiatives.
AI-driven scams are hitting Canadian employers’ bottom lines while leaving workers anxious and uncertain about whether they can spot the next attack, according to previous reports.
