Interactive reference

    The AI Exposure Map

    Every chart you have seen scores whole jobs. Jobs are the wrong unit.

    Exposure lives at task level. Every operating function, from People and Finance to Legal, Revenue Ops and IT, splits into work that automates and judgment that compounds. Select a function to see its tasks.

    The chart · eight functions

    The map asks the question. The task layer answers it.

    People & HR

    4 automate · 2 compound

    Employee queries and first-line adviceEmbed

    Grounded self-serve from your own policies, with citations

    9 /10

    Onboarding and offboarding adminTooling Enablement

    The classic monthly manual grind, run end to end

    9 /10

    Comp benchmarking and cycle modellingTooling Enablement

    Collection and modelling automate; the pay call does not

    8 /10

    ER case documentation and chronologyTooling Enablement

    Timelines and letter drafts. Verification is non-negotiable

    8 /10

    Difficult conversations and coachingHuman Enablement

    AI preps the brief; the room is yours

    2 /10

    Org design and workforce strategyAudit

    AI widens the option set; the bet is yours

    2 /10

    Curious how your own function scores? Take the Readiness Assessment, about ten minutes

    01

    Read the low scores first

    Tasks scoring 0 to 3 are where the human value sits: judgment calls, trust, hard conversations. They get more valuable as everything around them speeds up.

    02

    The high scores are capacity

    Tasks scoring 7 and above are not headcount cuts. They are the hours a team gets back to spend on the low-scoring column. Cutting the people who hold the judgment is how AI adoption fails.

    03

    Exposure is not adoption

    These scores are what current tooling can reliably do with proper setup and verification, not what your team does today. Most teams capture a fraction of this. That gap is the readiness question.

    Methodology: task inventories from operating engagements across seven sectors. Scores rate what frontier models plus workflow tooling reliably produce with proper context and a verification step, judged per task rather than per function. Argue with them, that is what they are for.

    Where is your company on this map?

    This is the generic version. Get one built for your actual functions and team: your tasks scored, mapped to a plan. No cost.

    You will get a short PDF within two working days, scored for your actual functions.

    Want this done properly, on your own process? Or prefer a number first? Either way, no cost to find out.