Guelph ordered to release redacted pet necropsy reports, while wildlife research data stays excluded

The University of Guelph must disclose redacted companion-animal necropsy information after Ontario’s privacy adjudicator rejected its claims of competitive harm and protected third-party information.

The October 2, 2026 ruling in Order PO-4918, Appeal PA22-00469 concerns nine sets of reports totalling 11,994 pages. Owners’ personal information and the animals’ medical histories are not part of the information ordered released.

The result contrasts with the same adjudicator’s August 25 decision in a separate Guelph appeal, Order PO-4889, Appeal PA22-00470. That ruling excluded a wildlife-health dataset from access law because of its established connection to university research.

Together, the decisions turn on the particular records and statutory grounds—not a blanket rule that all veterinary information is public or all university data is protected.

Almost 12,000 pages, examined through representative samples

The companion-animal request sought necropsy reports, including toxicology and rodenticide testing, with personal information removed. The university identified reports grouped into five canine categories and four feline categories.

Given their volume, Information and Privacy Commissioner adjudicator Alec Fadel examined one representative final report from each of the nine categories. The university did not object, and the requester had an opportunity to respond. His findings would guide the university’s treatment of the remaining records.

Guelph argued that disclosure could damage the Ontario Veterinary College’s economic interests and competitive position under section 18(1)(c) of the Freedom of Information and Protection of Privacy Act. It also relied on section 17(1)(b), concerning confidential third-party information whose disclosure could discourage similar information from being supplied.

Competitive harm needed more than a general concern

The university said the veterinary college occupied an unusual position: unlike private competitors, it was subject to freedom-of-information law. An affidavit from its dean said disclosure could undermine clients’ and referring veterinarians’ confidence, threaten fee-for-service revenue and affect veterinary education.

Fadel acknowledged that difference and accepted that confidentiality could matter to people choosing veterinary care. But the records before him were post-mortem reports, not ongoing clinical files connected to identifiable clients.

The university had already removed owners’ personal information, and the requester did not seek the animals’ medical histories. What remained was the veterinarians’ own observations, findings, diagnoses and test results. Fadel found no evidence that owners could be identified from the remaining information.

An institution did not have to prove harm would actually occur, he explained, but it needed detailed evidence that the risk was real rather than merely possible. General assertions that clients or referring veterinarians might lose confidence did not meet that standard here.

Nor did having private-sector competitors automatically exempt a university’s records. The legislature had made universities subject to access law despite their participation in competitive markets.

Diagnostic work was generated by the university

The third-party-information claim failed for a different reason. The exemption requires protected information supplied to the institution in confidence, together with a reasonable expectation of a specified harm. All requirements must be met.

Guelph argued that owners and organizations supplied the information by submitting animal specimens for examination. It relied on an earlier order recognizing that information embedded in a sample may, in some circumstances, be supplied with that sample.

Fadel distinguished information extracted from a submitted item from information generated through an institution’s own observations and expertise. The remaining information in these reports was the veterinary college’s diagnostic work, not information supplied to it for purposes of the exemption.

Because that requirement failed, he did not decide the other parts of the third-party test. With both claimed exemptions rejected, he also did not need to apply the public-interest override.

Wildlife dataset had a different legal foundation

The August decision concerned a 487-page spreadsheet of wildlife necropsy and toxicology information extracted from the Canadian Wildlife Health Cooperative’s database.

The requester disputed whether the spreadsheet answered the request for reports. Fadel accepted the university’s evidence that it contained the complete database information for the relevant specimens.

There, Guelph relied on section 65(8.1)(a), which excludes records associated with research conducted or proposed by university employees or associated people. Unlike an exemption for particular information, that exclusion removes qualifying records from the Act’s access scheme.

Fadel accepted evidence that the dataset was created as a core part of the National Wildlife Health Surveillance Program under a research agreement. Its research connection was established when the data were produced—not merely because researchers might use it later.

He rejected a rigid separation between surveillance and research: systematic monitoring can also generate research knowledge. But a university’s possession of a record, use of a standard form or employment of a veterinarian would not alone establish the exclusion. The specific research-program connection was decisive.

That appeal was dismissed. Because the wildlife dataset was excluded from the Act, the public-interest override could not apply to it.

Limited disclosure ordered by November 2

In the companion-animal case, Fadel directed disclosure by November 2, 2026, following the redactions in the examined samples and applying similar redactions to the remaining records. He retained jurisdiction for 30 days to address disputes over the sampling approach.

The decisions determine access to records. They do not establish that any particular animal was poisoned, identify an owner or reveal the results of the underlying testing.