ChatGPT
Claude
Perplexity
Gemini
GrokA custom GPT can start the conversation, but it cannot carry an admissions service alone
For a Canadian university or college, a custom GPT is a quick way to test whether applicants will use an assistant. It will not by itself give you consent records, a log you control, a link to your CRM or guarantees about where applicant data is processed. Those gaps matter more in Canada than the model's fluency does.
This guide is for admissions, recruitment and IT leaders deciding between a custom GPT, a build on the API and a dedicated admissions agent. It builds on our guide to AI chatbots for student recruitment.
What a custom GPT contains
A custom GPT is ChatGPT configured with your instructions, your documents and optional connections. A custom GPT combines instructions, attached knowledge documents and actions that call third-party services through APIs, so a link to your CRM needs an API that you build and host.
That means the chat is quick to set up and the integration is not. Writing an inquiry into your CRM needs authentication, error handling and monitoring, and someone in your organization must own it after the pilot.
Fit by job
| Admissions job | Custom GPT alone | Gap to close |
|---|---|---|
| Answering questions from public pages | Reasonable for a pilot | Keep sources current |
| Program requirements and prerequisites by province | Risky | Dated sources; decline when unsure |
| Application deadlines (OUAC, provincial centres, direct) | Risky | Verified calendar per channel |
| Residency, tuition category and international fees | Risky | Hand off to staff |
| Capturing an inquiry with consent | Not built in | Consent text and stored record |
| Writing to the CRM | Needs custom API work | Endpoint, authentication, alerts |
| Booking an open house | Needs custom API work | Calendar integration |
| Bilingual service (English and French) | Uneven | Tested answers in both languages |
Canadian applicants can apply through several routes. The Ontario Universities' Application Centre (OUAC) handles Ontario applications, while other provinces have their own centres and many institutions take direct applications. An assistant that blurs these routes sends people to the wrong place.
Accuracy across provinces and languages
Admission requirements differ by province, by program and by whether an applicant comes from the CEGEP system, a Grade 12 pathway or abroad. A model answers confidently in every case. Your control is retrieval from sources you maintain, a visible date, and a refusal rule.
If you serve French-speaking applicants, test the French answers separately, since a good English result does not guarantee one. Our guide to chatbot hallucinations and guardrails covers the safeguards in detail.
Privacy: PIPEDA, provincial law and Law 25
The Office of the Privacy Commissioner of Canada explains how PIPEDA governs personal information handled in commercial activities, while public institutions are usually covered by provincial public-sector statutes. Which regime applies to you depends on your institution type and province, so confirm with your privacy officer.
Quebec adds further duties. The Commission d'accès à l'information publishes guidance on Law 25, which covers consent, privacy impact assessments and the handling of personal information transferred outside Quebec. If you recruit in Quebec, treat a cross-border chatbot as a transfer to assess, not a detail.
An applicant may mention a disability, a visa issue or financial hardship in a chat. Plan for that, with a privacy notice that mentions the assistant and a clear answer about where the data is stored and who can read it.
Vendor terms and data location
According to OpenAI's API data documentation, data sent to the OpenAI API is not used to train or improve OpenAI models unless you explicitly opt in, and this describes the API only, not ChatGPT apps. The terms for a custom GPT inside ChatGPT are separate and depend on the plan, so check the exact plan and sharing setup before any applicant data is involved.
By default the same page says abuse-monitoring logs are retained for up to 30 days, unless longer retention is required by law. Whether processing happens inside Canada is a separate question that your privacy impact assessment must answer from the vendor's current documentation.
Logs and escalation
You need to see what the assistant said, when and from which source. You also need a rule for passing the conversation to a person, with the context, so the applicant does not repeat themselves. A staff-configured chat window rarely offers either in a form you can search and export.
Three routes beyond the pilot
Build on the API with retrieval. Your developers own the model calls, grounding, front end and log. You gain control and take on maintenance. OpenAI announced the removal of the Assistants API on 26 August 2026 and names the Responses API as its replacement, so new builds should target the Responses API.
Buy a dedicated admissions agent. Specialist products ship with consent capture, escalation, a CRM connector and reporting. Skolbot is one example, a website chatbot that answers prospect questions, qualifies inquiries, books appointments and pushes leads to the CRM, with a WhatsApp add-on.
Combine them. Use a custom GPT internally to draft replies for advisers and a dedicated agent on the public site.
| Question | Custom GPT | API build | Dedicated agent |
|---|---|---|---|
| Time to first pilot | Short | Medium | Short to medium |
| Control of data flow | Limited | High | Set by contract |
| CRM and booking | You build it | You build it | Usually included |
| Upkeep | Staff time | Your developers | Vendor |
| Best fit | Internal drafting | Engineering capacity | Public admissions service |
This is a planning aid, not measured data. For criteria, read how to choose an AI agent for student recruitment, and for build versus buy, SaaS, custom or open source.
A path for the next quarter
- Pilot internally on public pages and record each wrong answer.
- Compare the failures with the table above.
- Decide whether you can own an API build.
- If not, ask vendors to show consent capture, escalation, log export and data location.
- Complete a privacy impact assessment before applicants use the tool.
An illustrative workflow, not a measured result: a prospect asks in French about a program at midnight, the assistant answers from a dated page, offers a callback and records the request with consent in the CRM.
Questions to put to any vendor or internal builder
Whichever route you pick, the same questions expose the weak points. Ask them in writing and keep the answers with your procurement file.
Where is applicant data processed and stored, and which sub-processors touch it? Can transcripts be exported, searched and deleted on request? Does the vendor support a privacy impact assessment with the documents you need, including for transfers outside your province or outside Canada?
Ask how the assistant handles a question that depends on province of residence, such as tuition category or program prerequisites, and whether it hands those to staff. Ask how French and English answers are tested. Then run a trial: send twenty of your hardest real questions in both languages and review the answers with an admissions adviser before anyone external sees the tool.
If the builder cannot show the consent text, the escalation path and the log in a live demonstration, treat the pilot as a prototype and keep it away from applicants.
