AI and machine learning by Thomas McKinlay, Stefano Puntoni and Serkan SakaSeptember 9, 2026Illustration by Ana MorenoPostSummary. Trust is a central obstacle to realizing the potential of AI agents at work. Employees are unlikely to give agents meaningful autonomy if they are uncertain about their reliability, intentions, orLeer en españolLer em portuguêsPostImagine this scenario: You’re a manager testing out a new AI agent “assistant” for your team, and as you set it up it asks for permissions. Three options appear:PostRead more on AI and machine learning or related topics Generative AI, Leading teams, Teams, Innovation, Experimentation, Automation, Digital transformation, Organizational culture, Organizational change and Organizational transformation
To Adopt AI at Scale, Employees Need to Trust Agents
Trust is a central obstacle to realizing the potential of AI agents at work. Employees are unlikely to give agents meaningful autonomy if they are uncertain about their reliability, intentions, or ability to act safely. Research suggests adoption improves when organizations are explicit about an agent’s limitations, emphasize competence over friendliness, connect its recommendations to users’ broader goals, frame it as a helper rather than an autonomous authority, and preserve human control over consequential decisions. The objective is not blind trust but calibrated trust: employees should understand where agents are capable, where they are likely to fail, and when they remain in control. Without that confidence, organizations risk reducing sophisticated agents to glorified chatbots—and failing to capture their potential value.
Trust in agent reliability is the core adoption blocker for AI agents at scale; employees resist autonomy without certainty on agent intentions. Leaders must solve trust gaps to accelerate automation and redefine team structures around agent-assisted work.
