Right to disconnect: managing after-hours contact in practice

The right to disconnect does not stop an employer sending a message after hours. It protects an employee who refuses to monitor, read or respond to contact — or attempted contact — from the employer, or work-related contact from a third party, outside the employee's working hours, unless that refusal is unreasonable. The obligation created for employers is therefore indirect but real: an employer cannot treat a refusal that is reasonable as a performance or conduct failure, and cannot take adverse action against an employee for exercising the right.

That framing changes what an employer needs to build. The question is not whether to ban after-hours email. It is whether the organisation can explain, for each role, when contact outside working hours is expected, why it is expected, what the employee is paid or otherwise compensated for in relation to that expectation, and who decides that a matter warrants contact at all. Employers that can answer those questions role by role are in a position to assess a refusal on the statutory factors. Employers that cannot are left arguing about an expectation nobody wrote down.

This guide sets out the implementation work: role-by-role contact rules, alignment between availability expectations and pay, escalation channels that carry genuinely urgent matters, manager instructions, scenario assessment, workplace-level dispute handling, and the records that support a defensible position. For the design of digital channels themselves — conduct in those channels, monitoring and privacy limits, psychosocial risk in online work and preservation of digital evidence — see our guide to online workplace conduct and psychosocial risk. AWS is a workplace consultancy, not a law firm, and this is general information rather than legal advice.

The right applies to national-system employees under the Fair Work Act 2009 (Cth). It does not apply to independent contractors, and it should not be described as covering every worker or every public-sector employment relationship. Whether a particular workforce is in the national system, and whether a particular engagement is employment rather than contracting, are questions to settle before the rest of the analysis is worth doing.

It has applied to employees of non-small-business employers since 26 August 2024, and to employees of small-business employers since 26 August 2025. The right is in force for both cohorts; it is not a change still commencing.

The right is not a National Employment Standards entitlement. It operates as a workplace right, which means it is supported by the general protections in the Act (ss 340–341) rather than by the NES enforcement pathway. The practical consequence is that the risk attached to mishandling a refusal is adverse-action risk, and the Commission pathway for right-to-disconnect disputes has its own procedure, described further below. The Fair Work Ombudsman's summary of the right to disconnect and the Fair Work Commission's page on what the right to disconnect is are the two primary starting points for any internal briefing.

The protection covers contact or attempted contact from the employer, and work-related contact or attempted contact from a third party. Third parties may include clients, suppliers, other businesses or members of the public. An employer that gives client contact details to staff, or publishes a mobile number on a signature block, has created a contact channel it does not directly control but which is squarely work-related.

Attempted contact is covered as well as contact. A missed call, an unanswered message or a calendar invitation sent at 10pm may constitute attempted contact depending on the facts, and the expectation attached to it is what will be examined. Where the practical expectation is that the employee sees the attempt and responds, the fact that no conversation occurred does not remove the issue.

Contact from a colleague raises the same question in a different form. A peer in another business unit, working their own hours, may generate contact outside the recipient's working hours without any manager intending it. Cross-time-zone operations produce this routinely: a Perth team's ordinary afternoon overlaps the end of a Sydney colleague's day, and a Singapore or London counterpart may be at their desk when Australian staff are not. Time-zone difference is an operational fact, not an answer to the reasonableness question. The controls that help are practical — publishing each team's working hours in the calendar system, agreeing overlap windows for cross-border work, and instructing third parties and offshore teams where after-hours matters should be directed instead of to an individual.

• Map every channel through which work-related contact reaches employees outside their hours: work email, personal mobile, messaging platforms, ticketing systems, shared client inboxes and rostering apps.
• Identify which channels involve third parties and what those third parties have been told about after-hours contact.
• Record each team's working hours and time zone somewhere managers and cross-border colleagues can actually see them.

The operative right is framed by reference to the employee's working hours. The compensation factor in the reasonableness test refers separately to compensation for being available outside ordinary hours, or for working additional hours outside ordinary hours. Keeping those two expressions distinct matters when a matter is assessed: the question of whether contact fell outside working hours is not the same question as whether the employee is compensated in relation to hours outside ordinary hours.

The employee's working hours are established on the facts, and the documents that may evidence them include the contract of employment, the position description, a published roster, an approved flexible or individual flexibility arrangement, the applicable award or enterprise agreement, and the actual pattern of work recorded in time or system records. Where an employer wants a defensible position on when contact is inside or outside working hours, the practical step is to record the working pattern for the role in those documents and keep the record current as arrangements change.

Approved leave is a period in which an employee is not working. Employers should not assume that an employee on annual leave, personal leave, parental leave or long service leave carries a residual obligation to monitor a work channel. Nor should employers construct a standalone presumption that does not exist in the statute — leave is a highly relevant fact in assessing whether a refusal is reasonable, together with all the other factors, rather than a separate statutory rule.

A flexible working arrangement, an individual flexibility arrangement or another agreed arrangement may affect the factual position where it changes when the employee performs work; whether it does depends on its actual terms, which have to be read rather than assumed. Where an arrangement records a start at 6am and a finish at 2pm, contact at 5pm falls outside those hours even though it falls inside the team's. Where hours are split across the day, whether the gaps are working time depends on what was agreed. Any arrangement of this kind should be written down with the hours actually agreed, because it will be the document a manager relies on when deciding whether to send a message.

Section 333M sets out the matters that must be taken into account in determining whether a refusal is unreasonable. Each has to be considered; none of them is a shortcut, and the statute leaves room for other relevant matters as well.

The reason for the contact. What made the matter arise at that hour, and could it have waited until the employee's next working period? A genuine operational escalation and a manager clearing their own inbox at 9pm are different reasons, even if the message text is identical.

How the contact is made and the level of disruption it causes. A repeated call to a personal mobile at midnight is a different intrusion from an email that arrives silently and is read the next morning. Method and disruption are assessed together, and the employer controls both.

The extent to which the employee is compensated to remain available, or compensated for working additional hours outside ordinary hours. The statute provides that compensation for this factor includes non-monetary compensation. An availability or on-call allowance, a penalty or an overtime entitlement may be relevant; so may salary and non-monetary benefits, but only to the extent they actually compensate the employee for remaining available or for additional work outside ordinary hours. A salary label or an aggregate amount does not decide the issue. Compensation is a factor to be weighed, not a purchase of unrestricted availability, and it is a separate question from whether an annualised-wage or contractual set-off arrangement is valid under the applicable instrument.

The nature of the employee's role and their level of responsibility. A duty manager, an on-call engineer and a graduate analyst are not in the same position. Seniority is relevant, but it does not decide the question on its own and it should not be treated as a general answer for anyone above a certain classification.

The employee's personal circumstances, including family or caring responsibilities. This is a mandatory factor, not a discretionary courtesy. Where an employer knows an employee has caring commitments in the evening, that knowledge forms part of the assessment.

One rule operates differently from the rest. A refusal is unreasonable if the contact or attempted contact is required under a Commonwealth, State or Territory law. That is a specific statutory rule about legally required contact. It is not a general emergency exception, and it should not be described to managers as one. Urgency is highly relevant to the reason-for-contact factor, but urgency alone does not determine the outcome.

All 155 modern awards contain a right-to-disconnect term inserted following the Commission's award variation process, and the terms are not identical in their operation across every instrument. Award-specific provisions may deal with standby, recall to duty, on-call arrangements or contact about emergency roster changes. An employer assessing a scenario needs to read the applicable award, not a generic summary of the right. If there is any uncertainty about which instrument applies to a cohort, resolve that first — our guide to award interpretation and classification risks sets out the method.

Where an enterprise agreement contains a right-to-disconnect term that is more favourable to employees than the statutory position, that term continues to apply. Employers with agreements should identify whether such a term exists before writing any policy, because a policy cannot displace the more favourable agreement term.

Contracts and policies cannot contract out of the statutory right. A clause requiring an employee to be contactable at all times cannot displace the statutory right and does not determine whether a particular refusal is unreasonable; it may be another relevant matter alongside the mandatory factors, and its scope, any compensation attached to it, the terms of the applicable instrument and each statutory factor still require assessment. What contracts and policies can usefully do is record the employee's working hours for contact purposes, record separately any ordinary-hours and pay treatment relevant to compensation and to the applicable instrument, state a specific and limited availability expectation with the compensation attached to it, identify the escalation channel, and set out how the employee should raise a concern. Where an existing policy or contract is being changed to do this, the consultation, notice and consent boundaries in our guide to updating workplace policies apply to the change itself.

In roles that genuinely need after-hours coverage, make the coverage explicit and paid rather than assumed. That means a named on-call roster with defined periods, a stated expectation of what the on-call employee must do — answer within a set time, attend site, or simply be reachable — and an allowance, penalty or agreed arrangement that corresponds to that expectation under the applicable instrument.

An on-call label without a roster, without defined obligations and without corresponding pay does not do this work. Neither does an annualised salary described in general terms as covering additional hours; where an award or agreement governs the employee, any annualised or set-off arrangement has to satisfy the requirements of that instrument, and the employer needs to be able to show what the arrangement covers.

Rostered coverage also addresses the operational problem behind after-hours contact. If a matter needs a response at 11pm, someone should be rostered and paid to receive it, rather than a manager working through a contact list until someone answers. Coverage that is designed, staffed and funded is both an operational control and a clearer position if a refusal is later assessed.

• For each role, record the working hours, any availability expectation beyond them, the ordinary-hours and pay treatment that applies, and the compensation attached to the expectation.
• Check the availability expectation against the applicable award or enterprise agreement before it is communicated.
• Where no compensation attaches to an expectation, either fund it or remove it — an unfunded expectation leaves the employer without evidence on the compensation factor.

Employers need after-hours contact to work when it genuinely matters. The way to achieve that is to define, in advance, what qualifies — rather than relying on the sender's judgement at the moment they feel the pressure.

Escalation criteria should be written in operational terms a manager can apply at 10pm without advice: a safety incident, a system outage above a defined severity, a security incident, a regulatory notification with a fixed deadline, a client or site event of a specified type. Anything that does not meet the criteria waits. The criteria should be short enough to be remembered and specific enough that two managers applying them reach the same answer.

The responsible channel matters as much as the criteria. A single escalation path — the on-call number, the duty roster, the operations desk — means that a genuine matter reaches someone who is rostered for it, and that an employee who is not on call has no reason to be monitoring anything. Where the escalation path is "whoever the manager can reach", the organisation has no control at all over which employees are effectively expected to remain available.

Set the response expectation explicitly for each escalation tier: who must respond, within what time, and what happens if they do not. Then hold the escalation path to it. An escalation channel that is bypassed when it is inconvenient ceases to operate as a control.

Some after-hours contact is generated by managers working their own extended hours and clearing their queue when they have time. That is a legitimate way to work, and it does not have to create an expectation on anyone else. Delayed send and scheduled send in email and messaging platforms allow a manager to compose at 9pm and deliver at 8am. Quiet-hours settings in messaging and rostering apps suppress notifications outside defined windows. Both should be configured centrally and explained rather than left to individual preference.

Managers also need explicit instruction on what to do when they think something cannot wait: check the escalation criteria, use the escalation channel, contact the rostered person rather than the most responsive person, and record what happened. Where a manager decides to contact someone who is not rostered and not compensated for availability, they should be able to say why the matter met the criteria.

Delegation and backup coverage close the remaining gap. A single subject-matter expert who is the only person who can resolve a class of issue will be contacted after hours regardless of policy, because the alternative is an unresolved incident. The response to that is cross-training and a named backup, not a reminder about the right to disconnect.

Third-party instructions should be handled deliberately. Clients, suppliers and contractors who hold an employee's mobile number should be told, in the service documentation or contract, which channel to use outside business hours. Where an employee's direct number has been distributed widely, change the published contact point and tell the recipients where after-hours matters now go.

• Configure delayed send and quiet hours as defaults, and brief managers on when to override them.
• Give managers a one-page instruction: the criteria, the channel, the rostered contact, and the record to make.
• Name a backup for each single-point-of-failure role and fund the cross-training to make the backup real.
• Tell clients and suppliers where after-hours matters go, in writing, and update service documentation to match.

A position description that says the role "may require work outside standard hours from time to time" states no scope, no limit and no compensation, and settles nothing. A description that says the role includes participation in a one-week-in-four on-call roster, with a response expectation of thirty minutes and a stated allowance, describes an agreed availability arrangement that can be assessed.

The distinction to hold across the whole workforce is between agreed availability, which is defined, bounded and compensated, and assumed all-hours availability, which is none of those things. Where a senior role carries a genuine expectation of responsiveness to material matters, state that expectation clearly, with its limits and the compensation attached to it, rather than leaving it implied.

Review remuneration against the expectation at the same time. Where a role has acquired an after-hours load that was never priced, the options are to reduce the load through coverage design, or to price it through an allowance, penalty or reviewed salary arrangement consistent with the applicable instrument. Leaving an unpriced expectation in place produces both a right-to-disconnect exposure and a workload issue that carries its own psychosocial risk; our guide to psychosocial risk management covers the broader hazard-control analysis.

This matrix is an assessment tool, not a verdict table. Each scenario turns on the mandatory factors and on the particular facts, the applicable instrument and the arrangements in place. The value of working through it in advance is that it identifies the operational control and the record an employer would need before the scenario occurs.

This example is hypothetical and is included to illustrate the assessment. It is not a case, a decision or a regulator position.

A logistics employer runs a distribution centre with a rostered after-hours duty supervisor and a documented escalation path for equipment failures. At 1am a refrigeration unit fails. The duty supervisor follows the escalation criteria and calls the maintenance coordinator, who is on the on-call roster for that week and receives an on-call allowance under the applicable agreement. The coordinator answers, arranges a contractor and logs the incident.

The controls that make this defensible are visible before the incident: the escalation criteria define an equipment failure of this severity as a matter that cannot wait, the coordinator is rostered and paid for availability, the contact went through the designated channel to the designated person, and the incident log records what happened and why.

Change one fact and the position changes with it. If the coordinator had not been on the roster, had no availability arrangement and no compensation, and was called only because they answer their phone reliably, the assessment would run differently across several factors — the reason for contact would still be strong, but role, compensation, disruption and any personal circumstances would all need to be weighed, and the employer would be relying on an expectation it never defined or funded. The organisational fix is coverage design, not persuasion after the event.

This example is also hypothetical.

A national professional services firm has a Perth-based analyst working 8:30am to 5pm local time, and a Sydney-based engagement manager working ordinary eastern hours. The manager routinely sends task messages between 6pm and 7pm Sydney time, which lands mid-to-late afternoon in Perth, and follows up with messages at 8pm Sydney time — 6pm in Perth — expecting acknowledgement before the next morning. The analyst has caring responsibilities from 5:30pm and stops responding after finishing time. The manager raises responsiveness informally as a performance concern.

The employer has three problems, and none of them is the analyst's refusal. The team's working hours were never published, so the manager was operating on eastern-time assumptions. There is no availability arrangement or compensation attached to evening responsiveness for the analyst's role. And the refusal was treated as a performance issue before any assessment was made. On these facts — the hour, the absence of any operational reason, the analyst's level of responsibility and known caring responsibilities — the identified facts support the reasonableness of the refusal, but the employer must work through each mandatory factor and any other relevant matter, including the applicable instrument, before reaching a conclusion. Characterising the refusal as underperformance before completing that assessment is precisely where general-protections risk arises.

The remediation is practical. Publish each team member's working hours and time zone in the calendar system. Agree an overlap window for east-west work and schedule collaborative tasks inside it. Configure delayed send so the manager can work their own hours without generating an expectation. Withdraw the performance characterisation, deal with any genuine workload or deadline issue as a resourcing question, and record the decision so the file reflects what actually happened.

A disagreement about whether a refusal is reasonable must first be attempted to be resolved at workplace level before an application to the Commission under this pathway. That is a statutory requirement, and it is also the stage at which the facts, the role, the arrangements and the reason for the contact can be established while they are fresh — which is what any later assessment or application will turn on.

A practical internal process is short, defined and handled by someone other than the manager whose contact is in question. Hear the employee's account, establish the facts, apply the statutory factors on paper, decide, and communicate the outcome with the reasons. Where the assessment shows the expectation was unfunded, undefined or avoidable, fix the underlying arrangement rather than resolving the individual complaint in isolation.

The evidence worth holding is specific. Identify the contact or attempted contact: what was sent, by whom, through which channel, and at what time in the employee's time zone. Record the reason the matter arose then, the disruption caused, the employee's role and level of responsibility, the pay and availability arrangements applying to the role, any personal circumstances the employer knew about, the applicable award or enterprise agreement term, and what was done to resolve the matter at workplace level.

The instruction to managers should be precise. Do not take action against an employee because they exercised the right. Do not characterise or discipline the disputed refusal before the factor-by-factor assessment is complete. If genuinely unrelated performance or conduct action is contemplated during that period, keep its evidence, decision-maker and reasons separate and contemporaneous, and obtain legal advice where the position is not clear — because action connected to the exercise of the right is where general-protections risk sits.

• Nominate who handles a right-to-disconnect complaint, and make sure it is not the person who sent the contact.
• Use a standard assessment record that steps through each statutory factor, so decisions are consistent across managers.
• Close the loop on the underlying arrangement — roster, channel, allowance or role definition — not just the individual complaint.

Either an employer or an employee may apply to the Fair Work Commission about a right-to-disconnect dispute once workplace-level resolution has been attempted. The Commission's guidance on disputes about the right to disconnect sets out the procedure and the forms.

The Commission may make orders in this jurisdiction, including an order that an employee stop refusing contact, or an order that an employer stop taking certain action — for example, stop requiring an employee to monitor or respond to contact, or stop taking disciplinary or other action against an employee for a reasonable refusal. It may also deal with the dispute by other means, such as a conference, mediation or making a recommendation or expressing an opinion. Arbitration of the dispute beyond that requires the agreement of the parties; the Commission does not automatically arbitrate.

The general protections run alongside this. Because the right to disconnect is a workplace right, adverse action taken because an employee has exercised it can found a general protections claim, with the reverse onus that applies in that jurisdiction. That does not mean any discipline occurring after a refusal is unlawful — the question is the actual reason for the action, and a manager may have entirely unrelated and well-documented grounds. It does mean the decision record needs to show what the reasons were, made at the time, by someone who can explain them.

As at 6 August 2026, the Commission has deferred issuing written right-to-disconnect guidelines pending sufficient guiding case authority, as reflected in its major case page on the award variation process and Statement [2025] FWCFB 266. There is no settled body of decided test cases to design against. Employers should treat this as a currency-sensitive point and revisit their position when guidelines are issued or a material decision is handed down.

AWS works with employers on the implementation side of this: mapping contact channels and after-hours load role by role, aligning availability expectations with rosters and remuneration, designing escalation criteria and coverage, briefing and training managers, and building the assessment and record templates used when a refusal is questioned.

AWS is a workplace consultancy, not a law firm. Advice on the construction of an award or enterprise agreement term, on a contractual set-off arrangement, or on the conduct of a Commission or general-protections proceeding should be obtained from a qualified legal practitioner.