Skip to content
plugboard.appRequest a demo

Security and privacy

Automated decisions

What a school needs to describe Plugboard in its privacy policy under Australian Privacy Principles 1.7 and 1.8: which features decide something about a person, or do something substantially and directly related to a decision, what each one uses, and where a person decides.

From 10 December 2026, Australian Privacy Principles 1.7 and 1.8, added by the Privacy and Other Legislation Amendment Act 2024, apply to every APP entity, including non-government schools. In summary, they apply where:

  • the school has arranged for a computer program to make a decision, or to do something substantially and directly related to making one;
  • the decision could reasonably be expected to significantly affect a person’s rights or interests; and
  • personal information about that person is used in the program’s operation.

The school’s privacy policy must then say:

  • the kinds of personal information those programs use;
  • the kinds of decisions made solely by a program; and
  • the kinds of decisions where a program does something substantially and directly related to the decision.

APP 1.9 gives examples of decisions that may affect rights or interests: a decision made under a law, one that affects a person’s rights under a contract, agreement or arrangement, and one that affects their access to a significant service or support. Read the principles themselves before relying on this summary.

A “computer program” here is any software, not only AI. A fixed rule that works out what a family pays counts as much as a model does.

Government schools are generally covered by their state or territory privacy law rather than the APPs. The descriptions below are accurate for them too.

Plugboard is the school’s service provider. The school chooses which of these features it uses, so the school decides what its privacy policy says.

A feature is listed when all three of these hold:

  1. It uses personal information about a person.
  2. Software makes a decision about that person, or does something substantially and directly related to one.
  3. The decision could affect money a family is asked to pay, a person’s access to accounts, devices or the school site, the attention a staff member’s welfare receives, or how a staff member’s work is judged.

Features that only sort and route the desk’s own work are under considered and not listed, with the reason, so a school can take a different view.

FeatureWhat software doesSolely by software?How to turn it off
Repair cover and chargesWorks out which of warranty, AppleCare+, the school’s cover agreement or the school’s charge rule pays for a repair, and the most a family’s excess can beNo. A person proposes every charge and a second person approves itSettings, Statuses & fields: switch off Cover agreements or Damage charges
Routine requestsIssues a pre-approved loan, consumable or device restart it recognises in a ticketYes, when the school sets it to AutoSettings, AI, Routine requests
Access requestsRecognises a request for a group, licence or mailbox and works out who must approve itNo. A person grants every account changeSettings, AI, Access requests
Joiners, movers and leaversPlans account changes when the student system or directory shows a person joined, moved or leftNo. A person approves the plan; approved steps then run on their scheduleSettings, Access, or turn off the Directory module
Leavers in the student-system syncMarks a person inactive in Plugboard when they leave the student system’s rosterYesTurn off the student-system connection
FlowsWhatever a flow the school builds does, such as locking a device when a loan is overdueOnly for flows published fully automated, while the school allows itFlows, Flow controls, or each flow
Safeguarding flags and requester toneFlags a student’s ticket to staff when it uses words on the safeguarding list; on a staff member’s ticket, reads the tone, may raise the priority one step and may add a possible welfare concernThe priority rise, on Auto. A flag decides nothingSettings, AI, Requester tone and Welfare concern review. The word list stays on
Staff experience summaryScores how the desk has treated each staff member as a requester over 90 daysNo decision is made; the score can inform oneNot switchable; limited by permissions
Contractor check at sign-inHolds a contractor at the front office when their Working with Children Check is expired or not recordedYes, until a staff member sights the checkNot switchable

What it decides. When a student’s device is lodged for repair, Plugboard works out how the repair is paid for, in a fixed order: the manufacturer’s warranty, then AppleCare+, then the school’s cover agreement with its repair vendor, then the school’s own charge rule. Under a cover agreement it also works out whether the excess applies, who pays it (the school, the family or nobody) and whether the device has used up its claims for the period. When staff propose a charge, the amount starts from the agreement’s excess or the repair’s recorded cost, and a charge above the excess is refused.

Who decides. Rules work out the path. A model never decides cover, a charge or an amount. Nothing is charged automatically: a staff member proposes each charge and a different staff member approves it before the family can see it. The only exception is a school where nobody else could approve, where that person may approve their own proposal; the approval is then recorded as self-approved. A charge can be edited while it is proposed and waived, with a reason, until it is paid.

This is a decision where software does something substantially and directly related to the decision: it sets the path and caps the amount, and a person makes the decision.

Personal information used.

  • The student’s person type and year level, where a cover agreement is limited by them.
  • The device assigned to the student: model, asset type, warranty end date, AppleCare+ status, and the warranty answer from the manufacturer where a warranty connection is set up.
  • The repair type, and the number of claims already made on the device in the agreement’s period.
  • The cost recorded on the repair.

Turning it off. The rules only read records the school holds. A school controls them by what it switches on and records:

  • The answer to How are student devices provided? in the Setup wizard sets cover agreements and damage charges once. A school-owned 1:1 program, or a mixed fleet, gets both; families buying through a vendor portal, families bringing any device that meets a spec, and shared school devices only get neither.
  • Settings, Statuses & fields keeps Cover agreements and Damage charges as switches after that. With charges off, nothing can be charged to a family.
  • Settings, Cover agreements: with no active agreement, every repair that is not under warranty or AppleCare+ falls to the school’s charge rule, which a person applies.
  • Settings, Modules: turning off Repairs stops lodging, cover and charges.

The Likely repair outcome AI feature (Settings, AI) shows staff what a repair probably needs and costs from the school’s past repairs. It never sets cover or an amount.

Suggested wording.

When a student’s device is lodged for repair, our IT service desk software uses the device’s warranty and insurance status, the type of repair, the student’s year level and the number of claims already made on the device to work out whether the repair is covered and whether an excess applies. A member of staff then proposes any charge to a family, and a different member of staff approves it.

What it decides. When a ticket asks for something the school has marked as pre-approved in its request catalogue, such as a loan device, a consumable from stock or a device restart, Plugboard can fulfil it with nobody asked: it issues the loan in the requester’s name, issues the stock, or restarts the device. It does this only for items that change no account, group, licence or mailbox, never for a request that arrived by email, and only when the AI model is confident enough. Each fulfilment is listed for staff with an undo, except a restart, which cannot be taken back.

Who decides. Software alone, when the school sets Routine requests to Auto. On Shadow it records what it would have done and does nothing; on Suggest a person decides.

Personal information used.

  • The text of the requester’s ticket. On managed AI it is pseudonymised before it reaches the model.
  • The requester, or the one other person the ticket names, matched to the school’s roster.
  • The requester’s person type, which decides whether a catalogue item is offered to them.

Turning it off.

  • Settings, AI, Routine requests: Off, Shadow or Suggest.
  • Settings, Access: mark no item as pre-approved.

Suggested wording.

Our IT service desk software can recognise a request for a routine item, such as a loan device or a consumable, in a support ticket and issue it automatically where we have pre-approved that item. It uses the text of the request, the requester’s role and their existing loans.

A school that keeps Routine requests below Auto can leave this out.

What it decides. When a ticket asks for access to a group, a licence or a mailbox, Plugboard recognises which item in the school’s catalogue is meant and shows staff a card: whether the person already has it, whether a licence seat is free, and who must approve (the item’s owner, the person’s manager, the IT team, or nobody).

Who decides. A person, every time. Account, group, licence and mailbox changes always need a staff member to press the button and confirm their identity again. Software recognises the request and routes the approval, which is substantially and directly related to the decision.

Personal information used.

  • The text of the requester’s ticket, pseudonymised on managed AI.
  • The person’s directory account, group memberships and licences.
  • The person’s manager, from the directory, where the item needs the manager’s approval.
  • Seat counts, including seats held by people who have left.

Turning it off.

  • Settings, AI, Access requests: Off.
  • Settings, Access: the catalogue decides which items can be requested at all, by whom, and who approves them.

Suggested wording.

When a member of our community asks for access to a system, group or licence, our IT service desk software matches the request to our catalogue and works out who needs to approve it, using the request, their existing access and their manager. A member of staff grants or declines every request.

What it decides. When the student system or directory shows that a person has joined, changed year level, campus or role, or left, Plugboard builds a plan from the school’s templates: for example create an account, add it to groups and assign licences, or for a leaver, disable the account, reclaim licences and recover equipment. The Joiner and leaver plans AI feature can propose templates; it never runs anything.

Who decides. A person approves each plan, or a group of plans with exclusions, and confirms their identity again to do so. Approved steps then run on the timing the school set, for example disabling a leaver’s account a set number of days after they leave. Undoing an account creation disables the account rather than deleting it.

This is a decision where software does something substantially and directly related to the decision: it decides what the plan contains and when each step runs.

Personal information used.

  • From the student system or directory: enrolment or employment status, start and end dates, year level, tutor group, campus, department and position.
  • The person’s directory account, groups and licences.
  • Devices and loans recorded against the person.

Turning it off.

  • Settings, Access: turn off or edit each template.
  • Settings, Modules: turning off the Directory module stops plans.
  • Settings, AI, Joiner and leaver plans: Off stops AI proposals only.

Suggested wording.

When our student or staff records show that someone has joined, moved or left, our IT service desk software prepares the changes to their accounts, access and equipment, using their enrolment or employment details. A member of staff approves each set of changes before it runs.

What it decides. Each time Plugboard reads the roster from the school’s student system, anyone it previously read from that system who is no longer on the roster is marked inactive in Plugboard. Their records stay, but they can no longer sign in to the portal or appear as active. An empty roster changes nothing. The check is made per person type: where more than one in five active people of a type would be deactivated, nobody of that type is, the run says so, and the other types proceed.

Who decides. Software alone, following the school’s own student system. The school’s decision is the change it made in the student system.

Personal information used. The person’s identifier, person type and presence on the student system’s roster.

Turning it off. Turn off or remove the student-system connection in Settings, Connectors. A person who reappears on the roster is made active again at the next sync.

Suggested wording.

Our IT service desk software follows our student information system: when a student or staff member leaves our records, their access to the IT support portal ends automatically.

What it decides. Flows are automations the school builds. A flow starts on an event, such as a ticket arriving, a loan becoming overdue or a person leaving, and runs steps the school chose: for example sending a notice, locking a device through the school’s device management, or changing a directory account. A flow can include a condition written in words, which the Flow decisions AI feature answers.

Who decides. By default a person gives the go-ahead for every step that changes something or uses a privileged action. A flow runs every step with nobody asked only when all of these hold:

  • the school has switched on Allow fully automated flows in Flows, Flow controls, which needs two permissions and, where the school asks for it, a second person’s approval;
  • the flow was tested and published as fully automated by someone allowed to publish fully automated flows; and
  • each step stays within the permissions it was published with.

Some actions can never run fully automated. Accounts that hold a directory role, or are on the school’s protected list, are always refused.

Personal information used. Whatever the school’s flow reads: the record that started it (a ticket, loan, repair, device or person) and the details a step looks up. A condition answered by AI sees the text of that record, pseudonymised on managed AI.

Turning it off.

  • Turn off a flow from Flows.
  • Turn off Allow fully automated flows in Flows, Flow controls; every fully automated flow asks for go-aheads again at once.
  • Settings, AI, Flow decisions: Off.

Suggested wording. This depends on the flows a school runs. For a flow that acts on a person:

Our IT service desk software automatically [action, for example locks a school-owned device] when [event, for example a device loan is more than seven days overdue], using [information, for example the loan record and the device assigned to the student].

What it decides.

  • The safeguarding word list. When a student’s ticket uses words on the safeguarding list, such as words about self-harm, bullying or feeling unsafe, staff who can read internal notes are flagged, with the school’s safeguarding contact. In the portal the student is shown the school’s safeguarding message. This is a word match, not AI.
  • Requester tone. For tickets raised by staff, Plugboard reads the tone of the requester’s own messages (calm to distressed) and their urgency. With AI it is read by the model; without AI, by rules. On Suggest it proposes raising the ticket’s priority one step; on Auto it raises it.
  • Welfare concern review. On a staff member’s ticket, the AI can add a possible welfare concern for staff to review. It is never automatic.
  • Plugboard reads no tone and no welfare concern on a ticket raised by a student, a parent or a sender who matches nobody, with or without AI.

Who decides. A flag decides nothing: a person follows the school’s own process. A priority rise on Auto is made by software alone and is shown to staff, who can change it.

Personal information used. The text of the requester’s messages on the ticket, pseudonymised on managed AI, and the requester’s person type.

Turning it off.

  • Settings, AI, Requester tone: Off stops the AI and the rules reading of tone, and the priority rise.
  • Settings, AI, Welfare concern review: Off.
  • The safeguarding word list cannot be switched off. It is a safety net, and it only alerts staff.

Suggested wording.

Our IT service desk software alerts our staff when a student’s support request uses words that may indicate a risk to their safety or wellbeing, so that a person can follow our wellbeing procedures. The software does not make any decision about the student.

See sentiment & staff experience for how each reading works.

What it is. Each night Plugboard summarises how the desk has treated each staff member as a requester over the last 90 days: time to first reply, how often their tickets reopened, how often they had to chase, their satisfaction ratings, the share of their tickets that read as frustrated, and a 0 to 100 score from those numbers. With AI on, one sentence may be added for a person whose score is low, written from the numbers alone. Nothing like it exists for students or parents.

Who decides. No decision is made by software. The summary measures the service a staff member received, not their own work, but a school should not use it to judge a person, and should say so in its policy if it does.

Personal information used. The staff member’s tickets, reply times, satisfaction ratings and tone readings.

Turning it off. The score has no switch. Only staff who can read both reports and people can see it, within their campus. Turning off Requester tone stops the AI sentence and the tone readings the frustrated share is counted from.

Suggested wording. None needed while it is not used to decide anything about a staff member. A school that uses it in that way should describe it.

What it decides. When a contractor signs in at the front office or a visitor kiosk, Plugboard checks the Working with Children Check recorded for them. If it has expired or was never recorded, the visit is held: no badge is printed until a staff member sights the card and records it.

Who decides. Software holds the visit; a staff member clears it. The check is recorded by hand and never looked up in a state register.

Personal information used. The visitor type, and the check’s number, state and expiry date.

Turning it off. The hold cannot be switched off. It applies only to visitors recorded as contractors, and only where the school uses the visitor register.

Suggested wording.

Our visitor sign-in system holds a contractor at reception, without issuing a visitor badge, when no current Working with Children Check is recorded for them, until a member of staff has sighted it.

These features use personal information but, in our view, do not decide anything that significantly affects a person. A school may decide otherwise and list them.

FeatureWhat it doesWhy it is not listed
Triage and School tags (Settings, AI)Set a ticket’s category, priority and the school’s own yes/no labels; automatic on AutoOrders the desk’s work. Nobody is refused help, and staff can change every value
Ticket routingSends an unrouted ticket to a queue; department transfers always need staffAs above
Merge duplicates and Link similar ticketsMerges the same requester’s repeat report of the same issue within 48 hours; links othersJoins two reports of one problem; nothing about the requester changes
Ticket tidy-up and Incoming message reviewNotes whether a ticket is waiting on someone; sorts incoming emailNever resolves a ticket, discards a message or changes its destination
Loan remindersSends due and overdue reminders for loan devicesA notice, not a decision. Any consequence is decided by a person or by a flow
Sign-in protectionSlows or refuses repeated failed sign-insA temporary security control applied to everyone equally
Data retentionDeletes closed records older than the period the school setsApplies to all records by age, at the school’s choice
Portal article suggestionsShows published help articles while a requester typesRetrieval only; the requester can always lodge
Device fixesProposes a fix for a staff deviceEvery run needs a person’s approval

What Plugboard records about each decision

Section titled “What Plugboard records about each decision”
  • Every AI proposal, with its mode, reason and the model that made it, is in Settings, AI activity, which can be exported.
  • Every charge proposal, approval, waiver and self-approval, every access change, every plan approval and every flow step is in the audit log.
  • Erasing a person removes the tone and welfare readings and the staff summary about them, and a staff member’s personal information export includes their staff summary.