Employers and Workers Must Share Responsibility for Managing AI Risks
Both employers and employees need to invest in AI training and oversight to manage the technology's workplace risks, according to new research from the Transamerica Institute and Marsh McLennan Agency.
The findings reveal a significant gap between AI adoption and workforce readiness. Eighty-two percent of employers are currently using or planning to use AI to supplement their workforce, with 97% of large companies (500+ employees), 95% of medium companies (100-499 employees), and 79% of small companies (fewer than 100 employees) either deploying or considering the technology.
Yet 40% of surveyed HR and risk professionals cited a top concern: investment in AI and adoption without appropriate employee training and upskilling. The same proportion flagged limited knowledge or awareness of AI risks and mitigations.
The Mismatch Between Spending and Capability
Employers planning to expand AI use over the next three years face workforce challenges. Among large companies currently using AI, 85% plan to increase deployment within three years. Smaller companies show less aggressive expansion plans-51% of small companies and 71% of medium companies said they would expand.
The Transamerica survey found that 89% of employers using or planning to use AI expect workforce implications. Sixty-seven percent predict certain jobs will transform, 50% expect new roles to emerge, and 36% anticipate job elimination.
Marsh McLennan's report concluded that organizations face a "growing mismatch between AI spend and AI capability." Most companies expect employees to integrate AI into existing workflows rather than redesigning work itself, the report said. Without clear strategy, work redesign, and an AI-ready workforce, AI remains peripheral-improving isolated processes instead of transforming operations.
Cybersecurity and Misinformation Top Concerns
HR and risk professionals ranked training employees to identify AI-generated misinformation and disinformation as their top AI priority. Forty-one percent of respondents flagged this need.
Inadequate cyber threat literacy ranked as the leading "people risk" overall. More than 70% of surveyed organizations experienced at least one material third-party cybersecurity incident in the past year. Phishing and social engineering remain the most common entry points, with compromised login credentials leading to ransomware and data breaches.
As AI tools improve at mimicking reality, deepfakes become harder to identify. Employees need skills to recognize these threats across the workforce.
What Employees Want
Workers recognize the need to develop AI skills. Sixty-three percent of employees said they would accept a 10% pay cut in exchange for better AI and digital reskilling opportunities, according to Marsh McLennan's Global Talent Trends 2026 report.
Recommended Steps for HR Leaders
Marsh McLennan recommended that HR and risk management professionals treat digital literacy, judgment, and accountability as core risk controls. This requires shared ownership across risk, HR, and technology departments.
To manage cybersecurity risks tied to AI, employers should:
- Reframe cyber risk to include broader digital and technology risk
- Identify, assess, and recruit talent with strong cybersecurity skills
- Equip employees and external providers to maintain vigilance and operate safely under pressure
- Expand cyber scenarios within risk planning
- Treat workforce behavior, training, and culture as first-line cyber controls
- Strengthen security through controls while reducing high-risk exposures
Organizations should communicate their vision for AI integration to workers at all levels, involve them in AI design and implementation, and provide ongoing training for the future of work. Employers should create a "safe space" for employees to train, test, and master AI tools while protecting information privacy and security.
Learn more about AI for Human Resources or explore an AI Learning Path for CHROs to build organizational capability.
Your membership also unlocks:
