Survey finds startling disinterest in ethical use of AI among business execs, with only 35% saying their org tries to use AI in a transparent, accountable way
Context & Ripple Effects
This 2021 survey lands mid-arc in a decade-long gap between what executives say about AI ethics and what they practice. Back in 2018, IBM found about 60% of polled execs worried about explaining how AI uses data to meet regulatory standards — nearly double the 29% of 2016 — yet this later survey shows only 35% saying their organizations actually try to operate AI transparently and accountably.
The disinterest also cuts against public expectations documented in related coverage: a 2019 survey found 82% of Americans want AI carefully managed even though only 41% support its development. Meanwhile, business adoption had climbed to roughly half of firms by 2022 before plateauing, meaning a large installed base of AI is running inside organizations that largely lack accountable-use practices.
First-order effects
- Most organizations deploying AI have no stated commitment to transparency or accountability, leaving the majority of enterprise AI systems operating without internal ethical guardrails.
- The 35% figure exposes that the explainability concern executives expressed in IBM's earlier study on meeting regulatory standards did not translate into organizational practice — stated concern and actual behavior diverged.
Second-order effects
- Firms competing for customers and talent against this backdrop can differentiate on accountable AI use, since the public-trust deficit captured by the 82%-of-Americans-want-AI-carefully-managed survey makes trust a scarce asset.
- As adoption stalls near the 50–60% plateau McKinsey documented, vendors and consultancies gain a selling point for governance tooling aimed at the unaccountable majority.
Third-order effects
- If the pattern holds, accountability arrives through regulation rather than voluntary practice, forcing organizations that skipped governance to retrofit it under compliance pressure.
- Persistent low voluntary uptake positions operational AI governance as a baseline requirement rather than a differentiator, reshaping how enterprises budget for and audit AI deployments.
The trend: Enterprise AI adoption is outpacing voluntary ethical governance, pushing transparency and accountability toward becoming regulated mandates rather than optional commitments.