Trust Attitudes in the Use of Artificial Intelligence
-
KPMG & Univ. of Melbourne
- (48 000 workers across 47 countries) - Nov 2024 - Jan 2025
-
Some key findings
- 50 % present AI-generated content as their own
- 67 % intentionally use AI at work
- 69 % say their organization uses AI
- 54 % report increased efficiency, quality, and innovation
- 43 % report increased revenue-generating activity
- 28 % report increased workload stress and pressure
- Organizational readiness
- 60 % have an AI strategy in place
- 54 % have responsible‐use policies
- Tool sources & skills gap
- 70 % use publicly available free AI tools (non-enterprise-grade)
- 42 % feel they have the skills/knowledge to use AI appropriately
- 28 % have formal or informal AI training
- 52 % feel they can use AI tools effectively
- Risk behaviors
- ≈ 60 % (“three in five”) have seen colleagues misuse AI tools
- ≈ 66 % (“two in three”) rely on AI outputs without evaluation
- over 50 % have made mistakes in their work due to AI
- Organizational readiness
-
Implications & leadership gaps
- Trust gap exists more between employees and leaders than with AI itself
- Leadership deficits in providing tools, policies, training, guidelines, and clear communication
-
Shifting skill requirements
- From individual productivity “co-pilot” tools to design and integration of agentic systems
- Urgent need for training on building, provisioning, and managing AI agent teams
Earlier studies
- 75 % of knowledge workers using AI
- 78 % of those bring their own tools to work without informing colleagues or managers (Microsoft: “BYO AI”)
- 48 % of workers uncomfortable telling managers they use AI for common tasks (Earlier Slack)