Part of the AI Regulation News hub.
Canada has opened a public consultation on AI transparency covering synthetic content, chatbot disclosure, system information, incident tracking and AI agents
Nothing here binds anyone. The value is in what the paper concedes: the Government says the strategy commits it to work on transparency but does not predetermine whether any measure follows.
Bottom line: A consultation and a discussion paper. Neither binds anybody, neither creates an obligation, and the paper states plainly that whether further action is needed and what form it should take are questions the consultation will inform.
Who this affects: Compliance officers and general counsel at Canadian deployers of AI; AI developers selling into Canada; privacy lawyers; accountants and auditors advising on AI disclosure; anyone operating a customer-facing chatbot or an agentic purchasing tool in Canada.
Issue date: The consultation opened 23 July 2026 and closes 23 September 2026. The discussion paper page carries a date modified of 13 August 2026.
What changed: The Government has moved from the June 2026 AI for All strategy to a specific scoping exercise on transparency, and has named five candidate areas for action rather than one.
Analysis: The fifth area is the one to watch. Tracking the activity and interactions of AI agents is not a labelling problem, and Canada is scoping it in the same document as chatbot disclosure. The paper also asks a question most consultations leave implicit: where existing market responses and legal frameworks are already adequate.
Primary sources: Discussion paper: Enhancing trust in AI through increased transparency · Consultation page: Have your say on advancing AI transparency in Canada
- Instrument (EN)
- Enhancing trust in artificial intelligence through increased transparency (discussion paper) and the accompanying public consultation
- Authority
- Innovation, Science and Economic Development Canada (ISED)
- Jurisdiction
- Canada (federal)
- Status
- Open consultation
- Bindingness
- None. A discussion paper and a call for views. No legal obligation is created and no policy outcome is predetermined.
- Issue date / next deadline
- Open from 23 July 2026; closes 23 September 2026
- How to respond
- Anonymous survey, or comments by email to AIConsultations-ConsultationsIA@ised-isde.gc.ca. Submissions are treated as public documents and may be posted online.
- What follows
- The Government will review submissions and publish a What We Heard report
- Primary source
- https://ised-isde.canada.ca/site/ised/en/have-your-say-advancing-ai-transparency-canada/enhancing-trust-artificial-intelligence-through-increased-transparency
What is open and until when
The consultation page states a current status of open from 23 July to 23 September 2026. Responses go through an anonymous survey or by email to the ISED consultations address. The page warns that submissions will be considered public documents and may be posted online, and asks respondents not to include personal or confidential information.
After it closes the Government will review submissions and publish a What We Heard report. The page adds a disclosure that is worth noting on its own terms: depending on the volume of submissions, AI tools may be used to process the feedback received.
The discussion paper carries a date modified of 13 August 2026, later than the 23 July opening. The PDF filename on the page is labelled as updated.
The five areas
The paper scopes five areas where the Government could potentially take action. Detecting and identifying AI-generated content. Empowering individuals to know when they are interacting with an AI system. Improving the availability of consistent and understandable information about AI systems, including their development, capabilities and limitations. Enabling the tracking of serious incidents related to AI systems. Advancing ways to better track the activity and interactions of AI agents.
For each area the paper describes the transparency challenge, reviews current market and jurisdictional initiatives, and poses questions. It does not propose draft text, and it does not favour an instrument.
On instruments the conclusion is explicit: the Government wants feedback on the range of policy instruments available, naming voluntary codes, industry standards and legislation, and on how those instruments could interact with one another.
Why the AI interaction area is broader than chatbot labelling
The paper separates three kinds of encounter. Communicative interaction, where an AI system engages a person through conversation or voice and the question is whether the person knows the counterpart is a machine. Process interaction, where AI is used to assess a job application or process a claim and the person may never touch the system at all. And ambient influence, where AI shapes what people see through recommendation, which the paper concedes may not constitute an interaction at all.
That middle category is the one with the most exposure for employers and insurers, and it is framed as an open question rather than a proposal: whether people should be informed that AI played a role in a decision or process concerning them.
The AI agents area
The paper's agents section is descriptive. It names OpenAI's Operator inside ChatGPT for browsing, form filling and booking workflows, Google's Help me schedule feature in Gemini reading email and calendar context to create invites, OpenAI's instant checkout added in September 2025, Perplexity's PayPal-integrated equivalent, and Amazon's Alexa shopping feature completing purchases.
It attributes the growth in organisational agent use largely to open standards, naming the Model Context Protocol and the Agent2Agent protocol as having simplified how developers connect agents to data and tools and promoted interoperability between agents.
What the paper asks is how to better track the activity and interactions of agents. It proposes no mechanism.
Context the paper supplies
Adoption figures come from Statistics Canada. On the paper's account, second quarter 2026 data shows 19.2% of Canadian companies reported using AI to produce goods or deliver services over the preceding twelve months, up from 12.2% a year earlier and triple the 2024 proportion.
The paper situates the consultation against work already underway: Bill C-36, the Protecting Privacy and Consumer Data Act; Bill C-34, the Safe Social Media Act, creating obligations for social media services and chatbots; the passed Bill C-16 on non-consensual sexualized deepfakes and Bill C-25 on AI-enabled electoral misinformation; a $50 million expansion of the Canadian AI Safety Institute; and a Trusted AI Certification program.
The four cross-cutting questions at the end are the ones a submission should answer directly. They cover when transparency matters most and where action would be premature, how measures should account for small and medium-sized enterprises, how to ensure coherence with provincial, territorial and international frameworks, and how to design for continued technological change.
What we did not verify
We opened both the ISED consultation page and the full discussion paper in English and took every fact, figure and characterisation here from those two pages. The consultation page carries a date modified of 23 July 2026 and the discussion paper 13 August 2026.
We did not open the PDF version of the discussion paper, the survey instrument itself, the AI for All strategy document, or any of the bills named (C-16, C-25, C-34, C-36). The Statistics Canada adoption figures are reported as the paper states them; we did not open the underlying StatCan release. We did not read the French version and cannot speak to differences between the two.
We will not characterise any of this as a legal requirement. No obligation arises from a discussion paper, and the paper itself says the strategy does not predetermine outcomes. We also make no prediction about what the What We Heard report will contain or whether legislation will follow.
Treat the closing date as the actionable item and the five areas as a preview of scope. If you deploy a customer-facing chatbot, an agentic tool that transacts, or an AI system inside a hiring or claims process in Canada, the questions in this paper describe the disclosure surface a future Canadian instrument would most plausibly cover. Submissions are public, so write them accordingly.
Source File
Open the ISED consultation page and confirm the current status line reading open from July 23 to September 23, 2026, the five bulleted areas, and the note that AI tools may be used to process feedback. Then open the discussion paper and confirm its date modified of 2026-08-13 and the four cross-cutting questions in the conclusion.
While the strategy commits the Government to advancing work on AI transparency, it does not predetermine the outcomes. Whether further action is needed and what form any action should take are questions this consultation will inform. ยท ISED, Enhancing trust in artificial intelligence through increased transparency, discussion paper dated 13 August 2026
FAQ
Does this consultation impose any obligation on my organisation?
No. It is a discussion paper and a call for views. The paper states that the strategy does not predetermine outcomes and that whether further action is needed is a question the consultation will inform.
When does it close and how do I respond?
It is open from 23 July to 23 September 2026. ISED accepts responses through an anonymous survey or by email to AIConsultations-ConsultationsIA@ised-isde.gc.ca. The consultation page states that submissions will be considered public documents and may be posted online.
Which five areas does the paper cover?
Detecting and identifying AI-generated content; knowing when you are interacting with an AI system; consistent and understandable information about AI systems including development, capabilities and limitations; tracking of serious AI incidents; and better tracking of the activity and interactions of AI agents.
Does the paper propose a specific rule for labelling AI content?
No. It sets out the challenge and asks questions, including whether the goal is to inform Canadians whenever AI was involved in producing content or more narrowly to help identify AI-generated content that misrepresents reality. It notes there is disagreement about how much AI involvement counts.
Related briefings
Sponsored Training
Practical AI training for regulated professionals, built around verification, documentation and a defensible process. See the courses.