Wrong time, wrong message: Pamu's AI restructure costs $27k over flawed redeployment and consultation

The ERA ruled Pamu's AI shake-up unjustified: it "consulted" after deciding and muddled redeployment. The company must pay about $27k in damages and lost wages.

Categorized in: AI News Human Resources
Published on: Feb 17, 2026
Wrong time, wrong message: Pamu's AI restructure costs $27k over flawed redeployment and consultation

The process seemed thorough but responses came at the wrong time entirely

An AI-led restructure has cost Landcorp Farming Limited (Pamu) nearly $27,000 after the Employment Relations Authority found its redundancy process unjustified. On January 30, 2026, the Authority ruled that despite looking solid on paper, the company's consultation and redeployment steps fell short in critical ways.

The case centers on Aiga Faamanu Roache (Nu), the Accounts Payable Team Leader since 2016. After Pamu invested in automation to handle manual processing, management proposed changes in May 2024: remove Roache's team leader role, have AP officers report to the Chief Financial Controller, and add a cross-functional Transaction Specialist.

What actually happened

Roache returned from leave on June 18, 2024 and was shown the restructure proposal via a slide deck. She was given six days-until June 24-to provide feedback. Two days later, on June 26, the decision was emailed to her while she was on sick leave, with "responses" to her feedback attached to the same message.

The Authority was clear: consultation means you propose, listen, consider, then decide. Member Sarah Kennedy-Martin wrote that proper consultation "involves the statement of a proposal not yet finally decided on, listening to what others have to say, considering their response, and then deciding what will be done." Pamu's sequence was the reverse-responses arrived after the decision was effectively locked in.

The redeployment mistake that set the tone

At the first meeting, Roache was told she would need to apply for redeployment. Even the HR Business Partner later acknowledged hearing something to that effect. Under the employment agreement, suitable alternative roles should have been offered, not put into a competitive process.

This wasn't corrected in a way that landed. When Roache resigned on July 1, 2024, her email showed she still believed she had to apply for internal roles like Transaction Specialist or Accounts Payable Officer. That misunderstanding sat at the heart of the unjustified process finding.

Why that position? Why then?

The Authority also queried why the company chose to eliminate the occupied AP Team Leader role when the Accounts Receivable Team Leader role was vacant and covered by a fixed-term contractor. Pamu argued AI-driven efficiency mostly impacted AP, but there wasn't a clear link between processing gains and removing a supervisory role rather than a team member role.

There were also ongoing tensions between Roache and her manager that HR had previously supported-more than just "normal dynamics." That context mattered.

The outcome

Roache had received three months' salary as a goodwill payment on exit. Pamu argued that should reduce remedies. The Authority disagreed; it was voluntary and unsolicited. Pamu was ordered to pay $18,000 for humiliation and injury to feelings, plus $8,900.15 in lost wages covering seven weeks until Roache found new work.

Why this matters for HR

Automation doesn't relax your obligations; it raises the stakes. The business case still needs to be specific and defensible. Your consultation must be genuine. And the language you use around redeployment must align with contracts and law-small words, big consequences.

Practical steps HR can apply now

  • Sequence your consultation correctly: Present a proposal that's truly open. Invite feedback. Consider it. Then decide. Don't attach your "responses" to a final decision email.
  • Watch your words on redeployment: If the agreement requires offering suitable roles, do not ask the employee to "apply." Put offers in writing with clear terms, timeframes, and support.
  • Justify role selection with evidence: If automation impacts processing volume, explain why that leads to removing a leader vs. a team role. Use data, workflows, and risk assessments-be specific.
  • Respect timing realities: Six days can be short, especially if someone is sick or recently back from leave. Extend consultation where needed and acknowledge availability constraints.
  • Keep conflict out of the frame: If there are relationship issues, add independent oversight to decisions to protect fairness and the appearance of fairness.
  • Document the "consideration" step: Keep a record showing how you weighed each piece of feedback and what changed (or why not). This is the proof consultation actually happened.

Language that avoids trouble

  • Avoid: "You can apply for the alternative roles."
  • Use: "We are offering you the following suitable role(s) in line with your agreement…"
  • Avoid: "The decision has been made; here are our responses."
  • Use: "We're considering all feedback and will decide after we meet again on [date]."

AI-driven changes need extra clarity

  • Map the impact: Show where automation removes work (tasks, volumes, error rates) and how that translates to role changes. Don't assume "less processing" = "no supervisor."
  • Assess alternatives: Redeploy into vacancies first, restructure teams second, disestablish as a last step. Document each alternative considered.
  • Support redeployment: Offer training, phased transitions, and check-ins. Make the path simple to accept.
  • Pilot before cutting: Prove the automation gains over a defined period, then decide with real data, not projections alone.

Useful references

For HR upskilling on automation and role design

Looking to build capability around AI-enabled process change, redeployment, and workforce design? Curate practical learning paths for your team here:

Quick checklist before you press "send" on any restructure pack

  • Is the proposal truly open to change, with a clear feedback window?
  • Are redeployment obligations correctly framed as offers (if required), not applications?
  • Do you have evidence linking automation impacts to the specific roles you're changing?
  • Have you considered and documented viable alternatives, including vacancies?
  • Will your responses to feedback be issued before the final decision-not after?
  • Is there independent oversight if there's prior friction between parties?

Small process errors can turn into big costs. Get the timing right. Get the language right. And make sure consultation is real, not just a slide deck ritual.


Get Daily AI News

Your membership also unlocks:

700+ AI Courses
700+ Certifications
Personalized AI Learning Plan
6500+ AI Tools (no Ads)
Daily AI News by job industry (no Ads)