McKinsey 2026 AI Trust Maturity Survey: Progress Amid Agentic Shift

TLDR

AI adoption is accelerating into production and agentic systems, but RAI practices—especially strategy, governance, and controls for autonomous agents—are lagging. Organizations investing significantly in RAI report higher maturity and greater value realization, yet gaps in accountability, training, and agentic oversight persist.

Key Survey Insights

McKinsey’s survey (Dec 2025–Jan 2026, ~500 organizations with AI governance responsibility) uses a five-dimension RAI Maturity Model (strategy, risk management, data/technology, governance, and new agentic AI governance/controls). Maturity levels range from foundational to comprehensive/proactive.

State of RAI maturity:

• Average score improved to 2.3 (up from 2.0 previously).
• Only ~30% of organizations at level 3+ in strategy, governance, and agentic controls.
• Asia–Pacific leads; tech/media/telecom and financial services outperform other sectors.
• Significant investment in RAI strongly correlates with higher maturity and material EBIT impact.

Emerging risks/challenges:

• Security and risk concerns are the top barrier to scaling agentic AI (nearly 2/3 of respondents).
• Inaccuracy and cybersecurity remain top cited risks.
• Active mitigation lags awareness across most risk categories.
• AI incident frequency stable (~8%), but confidence in organizational response has declined.

Organizational responses:

• Knowledge/training gaps are the leading barrier to RAI implementation (~60%).
• Organizations with explicit RAI ownership/accountability achieve higher maturity.
• RAI is increasingly seen as a business enabler (value, efficiency, trust) rather than pure compliance.

Why this story matters

As AI systems move beyond chat to autonomous agents that act, trigger workflows, and interact at scale, weak governance creates real exposure. The survey underscores that trust is not optional infrastructure—it is foundational for safe scaling and capturing value. Organizations that close these gaps (especially agentic controls) will be better positioned; laggards risk slower adoption, incidents, and eroded stakeholder confidence.

Sources

• McKinsey: “State of AI trust in 2026: Shifting to the agentic era” (March 25, 2026). https://www.mckinsey.com/capabilities/tech-and-ai/our-insights/tech-forward/state-of-ai-trust-in-2026-shifting-to-the-agentic-era
• McKinsey AI Trust Maturity Model and survey methodology (based on ~500 organizations).
• Related McKinsey analysis on agentic AI and responsible practices.

Featured Image Alt Text

McKinsey chart or graphic illustrating 2026 AI Trust Maturity Survey results, showing average score of 2.3 with gaps in agentic governance

Tags

McKinsey, AI Trust Maturity, Responsible AI, Agentic AI, Governance, RAI, Survey