Canada has no overarching or sui generis database law; protection flows from copyright, common law trade secret principles, contracts, and privacy legislation, with the provinces playing a significant role. Under the Copyright Act, databases may be protected as compilations, but only the selection or arrangement of data is covered, and the Supreme Court in CCH Canadian v. Law Society of Upper Canada requires the exercise of 'skill and judgment' — mechanical or near-instantaneous compilation does not qualify, and purely AI-generated works are not protectable.
Trade secrets are protected primarily through the common law tort of breach of confidence, requiring that data be confidential, communicated in confidence, and misused by the recipient. A notable 2020 development added Section 391 to the Criminal Code, formally defining 'trade secret' and creating offenses for obtaining or communicating trade secrets by fraudulent means. Contract law remains flexible, but Ontario's 2021 prohibition on most employment non-compete agreements removed a traditional tool for protecting confidential data in employment relationships.
PIPEDA governs personal information federally, supplemented by provincial statutes in Alberta, British Columbia, and Québec, and courts have recognized privacy torts such as intrusion upon seclusion (Jones v. Tsige). Canada has no legislation addressing ownership or use of AI training data; as of 2024, training datasets, synthetic data, and AI-generated outputs fall under existing copyright, contract, and privacy laws — a regulatory gap that remains open: the proposed Artificial Intelligence and Data Act (AIDA, Bill C-27) died on the Order Paper at prorogation on 6 January 2025, and Bill C-36 (2026) revives privacy reform only, with no AI framework.
Key takeaways
- Database copyright in Canada demands human 'skill and judgment' in selection or arrangement, excluding mechanical and AI-generated compilations.
- Trade secret protection rests on the common law breach of confidence tort, now reinforced by criminal offenses under Criminal Code Section 391 (2020).
- Ontario's ban on most non-compete clauses means employers should rely on confidentiality agreements and NDAs to protect data in employment contexts.
- Data contracts must be drafted with PIPEDA and provincial privacy compliance in mind, including meaningful consent for AI uses of personal data.
- AI training data and outputs are governed only by existing copyright, contract, and privacy law; AIDA died at prorogation in January 2025 and no successor AI bill has been introduced.
Primary legal instruments
AI-assisted summary of the report, editorially reviewed before publication · AI transparency
Protection mechanisms
Report-derived analysis per mechanism. The report text is fixed; committee members with contributor access can revise the rating (with an audit trail) and add notes below each block.
Recent Developments
Live updates contributed by committee members and the AI research bot — separate from the fixed report content above.
Canadian International Trade Tribunal Initiates Inquiry into Automated Data Processing Software
The Canada Gazette announced a preliminary injury inquiry by the Canadian International Trade Tribunal (CITT) concerning automated data processing software. This inquiry, designated PI-2026-003, will assess potential injury to domestic industries. Such proceedings can highlight the economic value and competitive aspects of software and the data it processes, implicating unfair competition considerations.
Privacy Commissioner Submits Recommendations on Proposed Consumer-Driven Banking Regulations
The Office of the Privacy Commissioner of Canada (OPC) provided feedback to Finance Canada on the proposed Consumer-Driven Banking Regulations. The OPC supports the regime's goals for consumer privacy and data control, including provisions for data direction, breach reporting, and multi-factor authentication. The submission also suggests enhancements, such as clearer information for consumers, stricter limits on consent exceptions for public data, and robust security safeguards for participating entities.
Copyright Board publishes updated SOCAN tariffs for motion picture theatres
The Canada Gazette announced the publication of updated tariffs by the Copyright Board for the Society of Composers, Authors and Music Publishers of Canada (SOCAN). These tariffs specifically apply to motion picture theatres for the periods of 2017-2024 and 2025-2027. This update clarifies the licensing fees for the use of copyrighted musical works in a commercial setting.
Commons Heritage Committee urges opt-in consent and transparency for AI training on copyrighted works
In April 2026 the House of Commons Standing Committee on Canadian Heritage presented its report on the impacts of artificial intelligence on the creative industries, with 13 recommendations. These include requiring clear labelling of AI-generated content, ensuring the Copyright Act applies to AI outputs, imposing transparency obligations on AI developers about the copyrighted material in their training data, and establishing an opt-in consent requirement before protected works can be used to train AI systems. The report feeds into the government's pending decision on a text-and-data-mining framework following its 2023-2024 copyright consultation.
CanLII and Caseway AI settle landmark database-scraping copyright suit
On March 20, 2026 the Canadian Legal Information Institute announced it had resolved all matters in its British Columbia Supreme Court proceeding (VLC-S-S-247574) against AI startup Caseway. CanLII had alleged the bulk, systematic download of over 120 GB and about 3.5 million records from its curated case-law database to build a competing AI legal tool. The confidential settlement means Canadian courts did not rule on whether compiled public legal information attracts enforceable database rights, leaving the scope of protection for curated datasets untested.