Key AI risks
▹Employee misuse of AI without clear acceptable-use rules
▹Widening skill gaps across job families
▹Unfair AI use in hiring, evaluation, and promotion
▹Uneven adoption leaving some functions behind
Governed through the RMF Playbook
Each function pairs the outcome to reach for this domain (the Playbook's About) with the Playbook's suggested actions.
Govern
Define what acceptable AI use at work means and set the boundaries around AI in employment decisions.
Playbook · suggested actions
Set acceptable-use and AI-in-employment policies.
Map
Set role-based expectations for AI competence across every job family and level.
Playbook · suggested actions
Define role-based competency expectations by job family and level.
Measure
Assess whether staff actually hold the AI competencies their roles require.
Playbook · suggested actions
Assess staff competencies and track training completion.
Manage
Build competence through professional development, onboarding, and reskilling.
Playbook · suggested actions
Deliver development, onboarding, and reskilling to close gaps.
Worked case studies · the RMF Playbook applied
10 scenarios
Each step is a Playbook suggested action for that function; the evidence line is the transparency & documentation you keep.
AI-assisted resume screening in HR
Human Resources · staff hiring
HR wants to pilot AI-assisted resume screening to manage a high volume of staff applications.
Govern
Apply the employee acceptable-use policy and the AI-in-hiring policy.
Map
Set role-based competency expectations, classify the screener as consequential, and identify affected applicants.
Measure
Bias-test the screener and assess the competency of the staff who operate it.
Manage
Keep humans in the loop on every decision and integrate training into onboarding and reskilling.
Outcome — The screener assists but never rejects; trained staff make every final call.
Evidence: Hiring-tool bias test + training-completion records
Rolling out a role-based competency framework
HR · workforce development
The institution needs clear AI-competency expectations that differ by job family.
Govern
Adopt the role-based competency framework as HR policy.
Map
Define expectations for each job family and level.
Measure
Assess current competency against the framework.
Manage
Close gaps through targeted development and revisit the framework.
Outcome — Every role has clear, assessed AI-competency expectations.
Evidence: Competency framework + gap assessment
AI in performance management
HR · performance policy
Managers want to use AI to summarize and rate employee performance.
Govern
The AI-in-performance-management policy sets strict boundaries.
Map
Identify where AI may assist and where it may not decide.
Measure
Check for bias and accuracy in AI-generated summaries.
Manage
Keep managers accountable for ratings; AI only assists.
Outcome — AI supports performance writing without driving the rating.
Evidence: Policy sign-off + manager guidance
Acceptable-use onboarding for new staff
HR · onboarding
New employees need to know the AI rules from day one, not months later.
Govern
Integrate AI acceptable-use into new-employee onboarding.
Map
Define role-specific expectations at the point of hire.
Measure
Confirm completion and understanding of the rules.
Manage
Refresh onboarding content as policy evolves.
Outcome — New staff start with clear, role-specific AI expectations.
Evidence: Onboarding module + completion record
Reskilling a unit whose work AI automates
HR · workforce transition
A unit’s routine tasks are being automated and staff need new roles.
Govern
Workforce-transition guidance frames responsible reskilling.
Map
Identify affected roles and the competencies to build.
Measure
Assess reskilling progress against new-role requirements.
Manage
Deliver reskilling and redeploy staff into new work.
Outcome — Staff move into higher-value roles rather than being displaced.
Evidence: Reskilling plan + redeployment record
A manager using AI for evaluations
HR · manager practice
A manager quietly drafts evaluations with a consumer AI tool using staff data.
Govern
Acceptable-use rules bar sensitive staff data in unapproved tools.
Map
Identify the personal data the practice exposes.
Measure
Assess the exposure and the manager’s competency gap.
Manage
Provide an approved tool and manager training.
Outcome — Managers get safe tools and training instead of shadow practice.
Evidence: Acceptable-use guidance + training record
IT staff AI coding competency
IT · professional development
Developers use AI coding tools but competency and secure-use vary widely.
Govern
AI-assisted coding is framed as an employee-competency question.
Map
Define the secure-use competencies developers need.
Measure
Assess competency and secure-use practices.
Manage
Deliver development and pair it with Domain 5 code standards.
Outcome — Developers use AI coding tools competently and securely.
Evidence: Competency assessment + training tracker
A communications team using generative AI
Marketing & Communications · staff practice
Communications staff use generative AI for public content and need clear competency.
Govern
Acceptable-use and disclosure rules apply to staff-generated content.
Map
Define the competencies for responsible generative-AI use.
Measure
Assess whether staff can use it accurately and disclose it.
Manage
Train the team and connect to public-content standards.
Outcome — Communications staff use AI competently and transparently.
Evidence: Competency check + disclosure guidance
Finance staff using AI with sensitive data
Finance · staff practice
Finance staff want to use AI on budget and personnel-adjacent data.
Govern
Acceptable-use rules restrict sensitive financial data in AI tools.
Map
Identify which data is sensitive and off-limits.
Measure
Assess staff understanding and data-handling practice.
Manage
Provide approved tools and targeted training.
Outcome — Finance staff use AI without exposing sensitive data.
Evidence: Data-handling guidance + training completion
Measuring AI competency and credentialing
HR · assessment
The institution wants to verify — not assume — that staff hold required AI competencies.
Govern
Competency-assessment and credentialing standards apply.
Map
Define what each role must demonstrate.
Measure
Assess and credential staff against the standard.
Manage
Track completion and refresh credentials over time.
Outcome — Required AI competencies are verified and recorded, not assumed.
Evidence: Credentialing records + training tracker
Risk-tier examples
Assistive
AI grammar and style tools used by staff.
Operational
AI-assisted resume screening for staff hiring.
Consequential
AI systems making or recommending employment decisions.
Tools & artifacts
Role-based competency framework
Acceptable-use policy
Hiring-tool bias test
Training tracker
Key controls & instruments
Employee AI acceptable-use policy
Role-based AI competency framework
AI professional-development requirements
AI-literacy integration into onboarding