Right to Disconnect: Setting Work Boundaries in Tech

The law is narrower than the headlines. If you want to know what is the right to disconnect, start with the honest version: it is a patchwork of national rules, most of which protect you only when your refusal to respond is reasonable — and reasonableness is judged against your role, your seniority and your on-call terms. Australia’s version has applied to large employers since 26 August 2024 and to small-business employers since 26 August 2025 (Fair Work Commission). The EU has no directive at all.

That gap matters because the pressure is measurable. Microsoft’s telemetry, drawn from trillions of aggregated, anonymized Microsoft 365 signals, found the average worker receives 117 emails a day and 153 Teams messages per weekday, is interrupted every 2 minutes by a meeting, email or notification, and that meetings after 8 pm are up 16% year over year (Microsoft WorkLab). By 10 pm, 29% of active workers are back in their inbox.

So the durable protection is not statutory. It is a documented boundary contract: written response-time expectations, on-call terms in the offer letter, and evidence you can point to at review time. Here is how the rules read, and how to build one.

The short answer

Most “right to disconnect” laws protect a reasonable refusal, not an absolute one — your role and on-call terms decide the outcome.
Australia has an enforceable dispute path via the Fair Work Commission; the EU does not yet have a directive.
Ireland has a code of practice; Ontario requires a written policy, not a new right.
The United States has no federal right to disconnect — the question there is FLSA compensation, not boundaries.
The boundary that survives a performance review is the one written into your offer letter and calendar, not the one you assert verbally.

How This Guide Was Built

This is desk research, with no hands-on testing of any employer’s policy or any jurisdiction’s dispute process. The primary sources are the Fair Work Commission on Australia’s small-business commencement date, the European Parliament on the absence of an EU framework, the Workplace Relations Commission on Ireland’s code, ontario.ca on the Ontario policy requirement, and the US eCFR on FLSA hours-worked rules.

What was verified: the wording of each rule, its effective date, and the telemetry figures from Microsoft WorkLab and Gallup. What was not tested: any live dispute, any employer’s actual enforcement, and any negotiation outcome. Check your own contract, award or local rules before relying on any of it.

Last verified: September 2026.

What is the right to disconnect?

The right to disconnect is a legal or policy entitlement to refuse work-related contact outside your working hours without being penalised for it, and it exists because the boundary between work and rest has already collapsed. The telemetry is blunt: the average employee sends or receives more than 50 messages outside core business hours, and nearly 20% of employees who work on the weekend check email before noon on Saturday and Sunday (Microsoft WorkLab). That is not a personal discipline failure. It is an architecture problem, and it is expensive. Gallup’s global engagement update found low engagement cost the world economy approximately $10 trillion in lost productivity, about 9% of GDP (Gallup).

The rules exist because the always-on default is now the baseline, not the exception. Microsoft’s report also found that 48% of employees — and 52% of leaders — say their work feels chaotic and fragmented, and 1 in 3 say the pace of work over the past five years makes it impossible to keep up (Microsoft WorkLab).

What matters for you is that almost every version of this right is conditional. It is a shield against unreasonable contact, not a wall against all contact.

Does Australia’s right to disconnect survive the reasonableness test?

Yes, but only conditionally — Australia’s right lets employees refuse to monitor, read or respond to contact outside working hours unless that refusal is unreasonable, and for small-business employers it applies from 26 August 2025 (Fair Work Commission). For larger employers it has applied since 26 August 2024.

The carve-out is doing most of the work. “Unreasonable” is not defined by a stopwatch; it is assessed against the nature of your role, your level of responsibility, whether you are compensated for availability, and how the contact was made. For engineers, that means the same message can be protected for a mid-level IC and unprotected for a staff engineer who signed an on-call rotation.

Where on-call sits is the crux. If your contract, award or roster names you as on-call, the expectation to respond is usually built into the arrangement — and often into a loading or allowance. If it does not, an after-hours page is exactly the kind of contact the right was written for.

Disputes that cannot be resolved in the workplace can go to the Fair Work Commission (Fair Work Commission). That is a real escalation path, and it is also a slow one — which is why the written terms you negotiate before you need it matter more than the remedy you invoke after.

Check your own contract, award or local rules; nothing here is legal advice.

What do the EU, Ireland and Ontario actually require?

They require very different things, and the difference is the point — the EU has no directive, Ireland has a code, and Ontario requires a written policy rather than a new right. The European Parliament’s own legislative train states plainly that “there is currently no EU legal framework directly defining and regulating the right to switch off,” with the Working Time Directive (2003/88/EC) covering related rights only indirectly through minimum daily and weekly rest (European Parliament).

That said, the political direction is clear: on 21 January 2021 the European Parliament adopted a legislative-initiative report by 472 votes in favour, 126 against and 83 abstentions, calling on the Commission to propose a directive on minimum standards for the right to disconnect (European Parliament). A vote is not a directive, but it tells you where the floor is heading.

Jurisdiction	What the rule actually requires	What it means for you
Australia	Employees may refuse to monitor, read or respond to out-of-hours contact unless the refusal is unreasonable; unresolved disputes can go to the Fair Work Commission (Fair Work Commission)	Your protection depends on your role, seniority and on-call terms — get those in writing
EU	No legal framework directly defines the right to switch off; the Working Time Directive covers rest indirectly, and Parliament has called for a directive (European Parliament)	Your leverage is national law and your contract, not EU law
Ireland	The Workplace Relations Commission maintains an official Code of Practice on the right to disconnect (Workplace Relations Commission)	A code shapes expectations and evidence; it is not a standalone claim
Ontario	Employers with 25 or more employees on 1 January must have a written disconnecting-from-work policy in place before 1 March; the ESA defines disconnecting as not engaging in work-related communications but creates no new right (ontario.ca)	Ask for the policy in writing — the document is the deliverable
United States	No federal right to disconnect; after-hours contact is assessed under FLSA “hours worked” rules (29 CFR Part 785)	The question shifts from boundaries to whether you were working and paid

Is there a right to disconnect in the United States?

No — there is no federal right to disconnect in the United States, and the legal question is not whether your employer should contact you but whether that contact counts as compensable work under the Fair Labor Standards Act’s “hours worked” rules (29 CFR Part 785).

That reframing matters for engineers. A quick “can you take a look” Slack message is usually not the issue. A two-hour production incident at 11 pm is, and the analysis turns on whether you were suffered or permitted to work, how long, and whether you are exempt. For exempt salaried engineers, the compensation route is largely closed, which is why the boundary has to be negotiated rather than litigated.

So in the US, treat the boundary as a contract term, not a compliance claim. Your leverage is the offer letter, the on-call rotation document and your manager’s written expectations.

How do you set boundaries that survive a performance review?

You make them written, specific and reciprocal before you need them, because a boundary asserted in the moment reads as resistance while a boundary documented in advance reads as process. The goal is not to respond less. It is to make your availability a defined term rather than an assumption.

Define a written response-time SLA. Put it in your own words and share it: “I monitor Slack and email 09:00–18:00 local. Outside those hours I check once each morning before standup.” A stated SLA is a process; silence is a gap someone else fills.
Get on-call terms into the offer letter. Ask for the rotation cadence, the escalation path, the response window and any allowance in writing before you sign. If it is not in the letter, it is a favour, not a term.
Name the escalation path. Write it down and post it in the channel: “If this is a Sev1 outside hours, page the on-call engineer via PagerDuty. Slack DMs are not an escalation path.” This protects you and the person paging you.
Do notification hygiene. Turn off push for email and non-critical Slack channels; keep paging on. Microsoft’s telemetry found employees are interrupted every 2 minutes by a meeting, email or notification (Microsoft WorkLab) — you cannot reason your way out of a default you have not changed.
Block the calendar, not the intention. Put a recurring “Focus — no meetings” block and a hard stop at the end of your day. A visible block is a scheduling fact; a preference is a request.
Use the deferral script verbatim. “I’ll pick it up at 09:00 tomorrow — if it’s urgent before then, page the on-call engineer.” This acknowledges, defers and routes in one sentence, and it never says no.
Send a manager alignment note. After any boundary conversation, email a three-line summary: what you committed to, what you did not, and what you will escalate. “Confirming what we agreed: I’m on point 09:00–18:00, on-call per the rotation doc, and Sev1s go to the pager.” Undocumented agreements decay.
Separate availability from responsiveness. Being reachable is not the same as being obliged to reply. Say which one you are offering: “I’m reachable for Sev1s; I’m not monitoring for status updates.”
Renegotiate at review time, not in the incident. Bring evidence: the rotation doc, the SLA, the escalation path, and the incidents you handled outside hours. Ask for the terms to be formalised or the load to be redistributed.
Collect review-time evidence continuously. Keep a running note of after-hours work, pages answered and scope absorbed. Performance reviews reward what is visible, and after-hours work is structurally invisible.
Make the trade explicit when you say yes. “I can take this tonight — I’ll start at 10:00 tomorrow and shift the migration review.” A boundary that never costs anything is not a boundary; a trade is easier to approve than a refusal.
Escalate the pattern, not the incident. If after-hours contact is routine rather than exceptional, raise it as a staffing or rotation problem with the data attached. One late message is noise; a pattern is a resourcing decision.

Common mistakes

Treating the law as absolute. The right is conditional on reasonableness → read your contract, award and on-call terms before you rely on it.
Asserting a boundary verbally in the moment. It reads as resistance → send a written alignment note the same day.
Confusing on-call with always-on. A signed rotation is a term; an unprompted DM is not → point to the escalation path.
Negotiating after the incident. Leverage is highest before you sign → put the response window and allowance in the offer letter.
Keeping no record of after-hours work. Invisible work is unrewarded work → maintain a running log and bring it to review.

FAQ

Can my employer fire me for not answering after hours?

In Australia, an employee can refuse out-of-hours contact unless the refusal is unreasonable, and unresolved disputes can go to the Fair Work Commission (Fair Work Commission). In the United States there is no federal right to disconnect (29 CFR Part 785). Your contract, award and local rules govern — check them before you test the boundary.

Does the right to disconnect apply to on-call engineers?

Usually not in the way engineers hope. If your contract, award or roster names you as on-call, the expectation to respond is typically part of the arrangement rather than an unreasonable intrusion. Reasonableness turns on your role, seniority, the escalation path and whether you are compensated for availability. Get the rotation terms in writing.

What do I do if my manager ignores my boundary?

Escalate the pattern with evidence rather than repeating the request. Send a written summary of what you agreed, the after-hours volume, and the impact on delivery, then ask for a decision: formalise the terms, adjust the rotation, or redistribute scope. If your jurisdiction has a dispute path, such as the Fair Work Commission in Australia (Fair Work Commission), note it as a backstop.

Where to go next

Boundaries are one layer of a larger system. If you want the operating model rather than the legal one, start with how to build a work-life balance system that holds up. If you are already past the warning signs, read recovering from burnout after months of always-on expectations. For the tooling layer, see time-management tooling for engineers, and for the structural alternative, the four-day workweek in tech. Browse career tools and company deep dives to check how a specific employer handles on-call before you sign.

← Back to all posts