thirdbrain.

What's possible with AI in people and HR

Evidence current to mid-2026.

In people and HR, AI is ready today for the writing and admin, like drafting job ads, policies and induction, and summarising survey and exit feedback, always with your team reviewing. Hiring decisions, award pay and investigations stay with people. The tables below show exactly which is which.

Each task below sits in one of three bands: a strong fit today, worth a careful pilot, or keep with people for now.

A clear breakdown of where AI does and does not fit across the core tasks of a people and HR team. A practical starting point, not the last word.


Where AI is mature: strong fit today

Proven and available today. AI does the bulk of the work and your team reviews it.

FunctionThe job todayWith AI
Job ads and position descriptionsTurn a rough brief into an accurate, inclusive adAI drafts from the brief; your hiring manager checks pay and requirements and owns the words
Policies, handbooks and induction contentWrite and update policies and induction materialAI drafts and restructures; your HR manager checks every entitlement against current law
Interview questions, scorecards and schedulingBuild the questions and scoring guide, coordinate calendars and panelsAI drafts the questions and rubric and books the slots; your hiring manager edits and runs the interview
Reference-check templates and questionsBuild the referee question set, then conduct the check before the offerAI drafts the template and the questions to probe interview concerns; you conduct the check and own the call
Training and learning contentBuild courses, learning paths and quizzes from source materialAI drafts the modules and paths; a subject-matter expert checks the facts before staff see them
Survey and exit-interview analysisRead every comment, group the themes, summarise for leadershipAI clusters the themes and drafts the summary; your HR manager checks it and guards anonymity

Where AI is emerging: consider piloting with a human gate

Promising but not yet proven at this scale. AI assists and your team stays in the loop, so trial it on a contained scope first.

FunctionThe job todayWith AI
CV screening and shortlistingRead every application, filter, build a shortlistAI summarises and extracts; you decide, nobody is auto-rejected
New-starter paperwork and onboardingCollect tax, super, bank and contract details, set up accessAI captures and runs the workflow; your team checks right-to-work and pay setup before first pay
Employee questions (leave, policy, entitlements)Answer repetitive questions from policy and the systemAI handles logistics; for pay, leave or award questions it cites the source and routes to your HR team
Employment contracts and lettersDraft offers, variations and agreements from templatesAI drafts from the template; your HR manager or a lawyer checks and signs
Performance review and feedback draftingTurn notes and feedback into fair, specific review textAI drafts from real inputs; your manager owns the rating and the conversation
Workforce reportingPull headcount, turnover, absence and diversity into reportsAI assembles the report and commentary; your HR manager checks the figures and definitions
Interview note-takingTake notes in the interview, write them up against the rubricAI transcribes and drafts the notes, with consent; your hiring manager owns attribution and the assessment

Where AI is not ready or suitable today: keep with people

These stay with people, either because the tooling is not reliable enough or because controls and compliance rule it out.

  • Award interpretation and the pay calculation. Stays in payroll software built on a rules engine. Classification, gross-to-net, super and Single Touch Payroll (STP) are unforgiving, and underpayment is now a criminal offence. Payday Super commenced 1 July 2026, so super now aligns to each pay cycle and errors surface faster. AI can flag anomalies, not run the calc.
  • Hiring, ranking and reject decisions. Independent testing shows AI screening carries race and age bias, the employer carries the discrimination liability, and reading a candidate’s face or emotions is not valid. You decide, and nobody is filtered out unseen.
  • Workplace investigations and grievances. Assessing credibility, weighing evidence and making findings is judgement, bound by procedural fairness. AI can organise the documents, not reach the finding.
  • Redundancy selection and consultation. Who is made redundant, and a genuine consultation, are legal-judgement calls under the Fair Work Act. AI can draft and organise, never choose.
  • Attrition and flight-risk prediction. No reliable model at a small employer’s data scale, and it risks labelling people on proxies for age, gender or caring duties. The call stays with you.

Worth weighing: for any of these, the upfront setup (loading your policies, awards and templates, agreeing the review steps) is a one-off, separate from the ongoing effort, which on the strong-fit items is mostly review. The time that comes back goes to the conversations and the calls AI cannot make: the candidate who needs a real read, the employee with a grievance that needs procedural fairness, the pay query that needs a human to check the award clause.

One honest gap for performance drafting: an AI-generated review can read as fair and specific even when the manager’s input notes are thin. The line manager still needs to own the actual assessment, not just approve the AI’s version of it. The document matters, and the conversation with the employee matters more.

This guide sits on top of the things that stay the same whatever your function. See the ground rules

Questions

The questions leaders ask.

Will AI replace my HR team?

No, it changes what they spend time on. The drafting, the admin and the reading of survey and exit feedback shrink, and their judgement, their conversations and their reading of people matter more. The compliance-critical work, award pay, investigations and who gets hired or let go, stays firmly with your HR team. The goal is an HR function that does more with the same people, not fewer.

Is it safe to use AI on sensitive employee information?

It is safe when you control where the data goes and your team stays in the loop. HR data is some of the most sensitive a business holds, so the rule is simple: keep processing in Australia where you can, turn data retention and training off, put as little sensitive detail in as the task needs, and keep your team deciding anything that affects someone's job. Used that way AI drafts and summarises while the judgement stays human. One date worth noting: from 10 December 2026 the Privacy Act requires any business using AI in significant decisions about people, including candidate screening, to disclose that in its privacy policy. Worth checking your policy covers it before that date.

Where should we start?

Pick something repetitive and easy to check. Job ads, policies and induction content are a common first step: AI drafts, your team checks the facts and owns the words. Summarising open-text survey and exit feedback is the other strong one, because reading hundreds of comments by hand is the real chore and your HR manager still sense-checks the themes. Prove one before adding the next.

Book a discovery call
Do we still need HR and legal expertise?

Yes. The work shifts from drafting and reading to judgement and advice, and the calls that carry legal weight, award interpretation, classification, investigations, dismissals and redundancy, stay with your team. AI helps assemble the draft or the documents faster; it does not make the call, and under Australian law the employer carries the liability either way.

What can AI do in people and HR now that it could not six months ago?

Quite a bit, which is why this is dated. Through early 2026 the Australian HR platforms shipped assistants that draft job ads, policies and review comments and answer routine staff questions from your own documents, and survey tools made AI comment summaries the default. A year ago those were demos or roadmap; today they are usable, with your team still checking the output and owning the decisions.

How current is this, and what if the tools have moved on?

The evidence is current to mid-2026 and we refresh it as the tools and the rules change, which in HR they do often. Where a finding comes from an independent source we lead with it; where it comes from a vendor we treat it as marketing. If something has shifted since you read this, the fastest way to get the current picture for your own setup is a quick chat.

Book a discovery call

A first conversation

See how this applies in your world.

Book a confidential chat. Tell us where you see the opportunity, in people and hr or anywhere else, and we will agree a sensible next step. No deck, no pitch.