01
Observed evidence
Individual readiness is outrunning organizational support.
Only 26% of AI users in Microsoft’s 2026 survey said leadership was clearly and consistently aligned on AI.
Source-aware, not trend-chasing
Research signals
What is changing, what we can reasonably infer, and what teams should test next. Claims are linked to primary or first-party sources and labeled by type.
Reviewed July 24, 2026
Observed evidence describes published findings. Operating guidance translates established frameworks into a practical design move.
02
Observed evidence
AI-exposed work is increasing the value of human-intensive skills.
PwC reports that newly added tasks in AI-exposed roles are 2.5 times more likely to rely on empathy, judgment, and creativity.
03
Observed evidence
People expect AI-capable task share to keep rising.
Close to six in ten respondents to Anthropic’s user survey selected a higher band for the share of work AI could do in twelve months.
04
Operating guidance
Human-AI responsibilities should be documented.
NIST’s AI RMF playbook recommends defining the roles, responsibilities, proficiency, and oversight involved in human-AI configurations.
Three numbers worth holding lightly.
Aggregate signals can frame a question. They cannot predict a specific role, company, or person.
19%
of surveyed AI users were in the Frontier zone
View source ↗2×+
faster skills change in the most AI-exposed jobs
View source ↗39%
of key job skills expected to change by 2030
View source ↗Primary source shelf.
These sources ground the current role, skill, governance, and workforce guidance. Read them in context before making decisions.
O*NET Database
Occupation, task, skill, work-activity, and work-context foundations
Open source ↗U.S. Department of Labor AI Literacy Framework
Foundational AI literacy content and delivery principles
Open source ↗NIST AI Risk Management Framework
Governance, oversight, documentation, and review practices
Open source ↗World Economic Forum Future of Jobs
Employer expectations about workforce and skill change
Open source ↗Microsoft Work Trend Index
Organizational readiness and worker adoption signals
Open source ↗PwC Global AI Jobs Barometer
Job, skill, productivity, and wage patterns in AI-exposed work
Open source ↗Interpretation boundary
Do not turn a trend into a verdict.
A role contains different tasks, people have different contexts, and technology changes. Use research to form better questions, then test the work locally with affected people.