← All models

9. AI in People Work

Now, next, later

Sequencing AI in HR

Use when

Building a credible roadmap instead of a vendor shopping spree

Now, next, later, shown as a sequence of 3 parts: Now, Next, Later.
  1. 1

    Now

    0-6 mo

    Admin relief with zero decision content: drafting, tagging, summarizing, theming anonymized text; ride tools already governed in-house

  2. 2

    Next

    6-18 mo

    Foundations and human-in-the-loop assists: skills taxonomy, data packs for talent processes, personalized learning

  3. 3

    Later

    18+ mo

    High-sensitivity plays after trust is earned: mobility matching, any predictive modelling, always heavy consultation

“Sequencing is trust management: visible harmless wins first.”

How to use it

Sequencing is trust management: visible harmless wins first, foundations second, sensitive plays only after the organization has watched AI behave for a year. The discipline that makes it credible is publishing the criteria for graduation between horizons and honouring the never-AI list throughout.

Card details

Model type
Sequence
Origin status
Synthesis
Content last reviewed
28 July 2026

Origin

Synthesis

Now, next and later is a common roadmap convention. What is specific here is the HR sequencing logic and the requirement to publish the criteria for graduating between horizons. Assembled by Harsimran Kaur Kapoor, 2026.

Neighbouring models

AI-fy, augment, keep humanGovernance gates for AI touching employees
NextCompeting commitments (Immunity to Change)→

Was this useful?