Access to Information Act, R.S.C. 1983, c. A-1

The Access to Information Act came into force on July 1, 1983, on a principle Parliament had debated for more than a decade: that citizens of a democracy have a right to know what their government is doing, subject to specific and limited exemptions. It was meant to be a structural piece of democratic machinery — the mechanism by which the public could turn the lights on.

Forty-plus years later, the Act is the primary instrument women's advocates, journalists, and concerned citizens have used to ask one straightforward question of Correctional Service Canada: how many male inmates have been transferred to women's institutions under the 2017 self-identification policy, and what has happened to the women in those institutions since?

The answers have come back — when they have come back at all — as heavily redacted pages, as assertions that no responsive records exist, and as months or years of delay that have outlasted the patience of the requesters.

What the Act Actually Does

The Access to Information Act codifies a two-part principle:

1. The right of access. Canadian citizens, permanent residents, and any corporation or individual present in Canada have a statutory right to request records under the control of a federal government institution, in any format those records exist.

2. The duty to disclose. The government institution receiving a request has a statutory duty to provide those records, subject to a closed list of exemptions and a limited time period — typically 30 days, with extensions available in defined circumstances.

The statute binds more than 260 federal government institutions, including Correctional Service Canada, Public Safety Canada, the Canadian Human Rights Commission, Statistics Canada, and every department and agency of the federal government. It does not bind the provinces, nor the courts, nor — until the 2019 reforms — the offices of the Prime Minister and Ministers.

The Act's purpose is stated in section 2(1): to provide a right of access to information in records under the control of a government institution, on the principle that government information should be available to the public. Section 2(2) reinforces that the Act is intended to complement, not replace, existing procedures for public access.

Or at least, that is how the statute reads. How it has operated in practice — particularly in response to requests about federal corrections and gender identity policy — is another matter.

The Exemption Framework

The practical question in almost every Access to Information request is not whether the requester has a right of access. They almost always do. The practical question is whether the requested records fall within one of the Act's enumerated exemptions.

The exemptions are specific:

• Section 13 — Information received in confidence from other governments

• Section 15 — International affairs, defence, and security

• Section 16 — Law enforcement and investigations

• Section 17 — Safety of individuals

• Section 19 — Personal information about individuals other than the requester

• Section 21 — Operations of government (advice, recommendations, consultations)

• Section 23 — Solicitor-client privilege

• Section 24 — Statutory prohibitions on disclosure elsewhere in federal law

Some exemptions are mandatory — the institution must refuse disclosure. Others are discretionary — the institution may refuse. For most advocacy-relevant requests about CSC's gender-diverse offender policy, the decisive exemptions have been sections 19 and 21.

Section 19 and the Personal Information Shield

Section 19(1) requires a government institution to refuse to disclose any record requested under the Act that contains personal information. Personal information is defined by reference to the Privacy Act — effectively, any information about an identifiable individual.

In the context of CSC data on trans-identified inmates, section 19 has been applied to withhold:

• The number of male inmates transferred to specific women's institutions

• The offence histories of transferred inmates, where release could identify them

• Incident reports involving named institutions

• Aggregate data sets where the small population size — fewer than five individuals in some institutions — means that even anonymized data could theoretically identify individuals

The result is a regime in which a request for "how many" produces a redaction for each "how many," on the theory that an individual's placement in a particular institution is personal information about that individual.

The public interest exception

Section 19(2) creates exceptions: disclosure may proceed where the individual consents, where the information is already publicly available, or where disclosure is in the public interest. CSC has declined, with rare exception, to treat public interest in the safety of incarcerated women as a sufficient basis for disclosure under section 19(2).

This is a choice, not a requirement of the Act. The Information Commissioner can recommend — and since 2019, can order — disclosure in the public interest. But the underlying cultural posture of the institution, that individual placement information is categorically off-limits regardless of aggregation, has not changed.

Section 21 and the "Advice to Government" Shield

Section 21 allows government institutions to withhold records containing advice, recommendations, or consultations among government officials — effectively, the deliberative material that went into a policy decision.

For requests seeking the policy rationale behind CSC's December 2017 decision to issue Interim Policy Bulletin 584, or the rationale behind Commissioner's Directive 100 in May 2022, section 21 is the shield CSC has relied on.

This matters because the core accountability question of both instruments is procedural as much as substantive. Who made the decision? On what evidence? With what consultation? Subject to what review? A section 21 exemption makes those questions unanswerable from outside the institution.

Successive Information Commissioners have noted that section 21 has been interpreted expansively over time, shielding not just active policy advice but also retrospective briefing materials, internal consultation records, and pre-decisional analysis that elsewhere would be treated as routine disclosure. The practical effect, in the specific case of CSC's gender-diverse offender policy, is that the public knows what the policy says, but not how it came to be.

The Review Mechanism

A requester dissatisfied with an institution's response can file a complaint with the Office of the Information Commissioner under section 30 of the Act. The Commissioner has the power to investigate complaints, recommend disclosure, and — since the 2019 amendments under Bill C-58 — issue binding orders.

The Office describes its mandate as follows:

The Information Commissioner is the first level of independent review of government decisions relating to requests for access to information under the Access to Information Act.

In practice, the Commissioner's intervention is a long process. Average complaint resolution times have exceeded two years. The Office has reported a persistent backlog, and the institutions under the greatest advocacy scrutiny — CSC, Public Safety Canada, the RCMP — are among the most frequent subjects of complaint.

The Commissioner's order-making power, introduced in 2019, was intended to address the limits of recommendation-based resolution. Early evidence suggests it has made a difference at the margins. It has not substantially changed the pattern of delay, redaction, and claims of "no responsive records" that characterizes CSC's handling of requests about its gender-diverse offender policy.

Bill C-58 and What the 2019 Reforms Changed

Access to Information received its most significant statutory update in 2019 with the passage of Bill C-58, which came into force on a staggered schedule between 2019 and 2020. The reform addressed long-standing criticisms of the Act, but not in ways that have substantially benefited requesters seeking data about gender identity and corrections.

Key changes included:

• Order-making power for the Information Commissioner — binding orders for disclosure, not merely recommendations

• Proactive publication requirements — ministerial mandate letters, briefing packages for new ministers, and travel and hospitality expenses are now published automatically

• Elimination of application fees — the $5 fee for making a request was removed

• Extended coverage — the Prime Minister's Office, Ministers' Offices, and administrative institutions supporting Parliament became subject to proactive disclosure, though not to the request-and-review regime

The reforms did not:

• Expand the definition of "information" that can be requested

• Narrow the exemption categories

• Alter the application of sections 19 or 21 in ways that would affect CSC's disclosure patterns

• Impose statutory deadlines for the Information Commissioner's complaint investigations

For women's advocacy seeking CSC data, Bill C-58 did not change the landscape.

What ATIA Requests Have Actually Produced

The figures that now appear in caWsbar's Charter challenge — 31% of trans-identified male offenders with sexual assault convictions, 55% of their victims women, 94% of assaults committed before the offender identified as transgender, 44% with previous sexual assault convictions — came from a combination of Access to Information requests and CSC's own published research.

These figures did not come quickly. They did not come completely. And they did not come without litigation of the request process itself.

The seven-step pattern

The pattern has been:

1. A request is filed for specific data on trans-identified inmates in women's institutions

2. CSC invokes extensions under section 9 — typically 30 to 90 additional days

3. A partial response is eventually issued, with section 19 redactions covering most specifics

4. The requester complains to the Information Commissioner under section 30

5. A prolonged investigation produces a recommendation (or, post-2019, an order) for additional disclosure

6. CSC releases additional material — often still heavily redacted — on the Commissioner's direction

7. The cycle repeats for follow-up questions

For an advocacy organization operating on volunteer labour and limited resources, the request process is itself a filter. The questions that get asked are the questions requesters can afford to pursue. The questions that do not get asked — or do not get pursued to the Commissioner — are effectively invisible to the public record.

The "No Records Exist" Problem

A consistent and troubling response pattern in ATIA requests to CSC on gender-diverse offender data has been the assertion that no records exist responsive to the request.

This response is, on its face, surprising. CSC maintains offender records as a statutory duty under the Corrections and Conditional Release Act. Offender placement decisions are recorded. Incident reports are filed. Policy decisions are documented. The institutional production of records is extensive and ongoing.

Three explanations have been offered for "no records exist" responses:

1. The specific record requested does not exist — the institution has not compiled the data in the form the requester has asked for

2. The records exist but are held by another institution — the request should have been directed elsewhere

3. The records exist but are not captured by the request as worded — technical non-compliance

The first explanation is the most significant for advocacy purposes. If CSC has not compiled statistics on the number of male-on-female sexual assaults in women's institutions by the transgender-identification status of the perpetrator, then that data simply does not exist in CSC's records, and there is nothing to release.

The Office of the Correctional Investigator has repeatedly recommended that CSC improve its data collection practices with respect to gender-diverse offenders and outcomes in women's institutions. CSC has not, to date, implemented those recommendations in a way that produces releasable aggregate data.

The Oversight Paradox

Democratic accountability depends on a feedback loop: government acts, the public observes, the public holds the government to account.

Access to Information is one of the structural mechanisms that makes this feedback loop possible. When it works, citizens and organizations can observe government action, identify problems, and bring those problems to public and political attention.

When it does not work — when exemptions are applied expansively, when redactions obscure rather than protect, when complaint investigations take years, when "no records exist" forecloses inquiry — the feedback loop is broken. The government continues to act. The public continues to pay. But the specific evidence needed to hold the government accountable remains inside the government.

This is the paradox the women's advocacy movement has encountered in Canadian federal corrections. The policy that put male-bodied inmates in women's prisons was not a secret. The existence of Interim Policy Bulletin 584 was public. Commissioner's Directive 100 was published on CSC's website in 2022. What was not public — and what remains not public — is the granular operational data about how the policy has worked in practice: how many transfers, to which institutions, of inmates with what offence profiles, producing what incidents, against what backdrop of staff and inmate complaints.

Without that data, the Charter challenge filed on April 7, 2025, must proceed on the basis of what has been disclosed under protest, pieced together from partial ATIA responses, Correctional Investigator reports, and the testimony of women willing to speak publicly. That is a harder case to make than a case built on complete records. And the harder case is, in significant part, a consequence of the Access to Information Act functioning as a filter rather than as a window.

The Charter Challenge as a Transparency Demand

The Charter challenge filed by caWsbar and the Justice Centre for Constitutional Freedoms on April 7, 2025, is not only a substantive challenge to cross-sex placement policy. It is also, implicitly, a demand for transparency.

Litigation processes require disclosure. Discovery in a Charter action will require CSC to produce records that ATIA requests have not successfully obtained. The court may compel production of documents and data that institutional discretion has declined to release. The adversarial process — examination of CSC officials, cross-examination of policy decision-makers, production of internal deliberation records — will be able to do what the administrative ATIA process has, on this subject, not been able to do.

This should not be necessary. Access to Information was designed precisely so that major policy questions could be understood by the public without the need for constitutional litigation. On the subject of male inmates in women's federal prisons, that design has not held.

What Happens Next

Three paths forward exist for strengthening access to information on this policy area:

1. Statutory amendment. Parliament could amend the Access to Information Act to narrow sections 19 and 21 in ways that permit greater disclosure of aggregate institutional data, particularly on matters of public safety and human rights. This has been repeatedly recommended by successive Information Commissioners and has been repeatedly deferred.

2. Policy change at CSC. The Commissioner of CSC could, as a matter of administrative practice, adopt a presumption of disclosure for aggregate data about gender-diverse offender transfers, consistent with the Act's underlying principle. This would not require statutory change.

3. Charter litigation outcomes. A successful Charter challenge could produce court-ordered disclosure, setting a precedent that would be difficult for CSC to resist in future requests. A successful challenge might also prompt statutory reform in response to the transparency failures exposed at trial.

The first path requires legislative will. The second requires administrative will. The third requires, and is building, the coalition of incarcerated women, advocacy organizations, and legal counsel willing to do the work.

Until one of these paths yields, the Access to Information Act will continue to be — for the specific purpose of asking what Canadian federal corrections does in the name of the women it holds in its custody — an instrument of partial answers.