ChatGPT
Claude
Perplexity
Gemini
GrokA custom GPT is a useful pilot for an Australian admissions team, not a full enquiry service
A custom GPT can answer common questions within a day of setup. It will not, by itself, record consent, keep a log you control, write to your CRM or know that a late-offer round has just opened. Those gaps decide whether a pilot can safely face applicants.
This guide is for admissions, marketing and IT leads at Australian universities and higher education providers choosing 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 you are configuring
A custom GPT is ChatGPT with your instructions, 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.
The conversation is therefore the easy part. Capturing an enquiry in your CRM needs authentication, error handling and monitoring, and a named person has to own that work once the pilot ends.
Where it helps and where it stops
| Admissions job | Custom GPT alone | Gap to close |
|---|---|---|
| Answering questions from public course pages | Reasonable for a pilot | Keep sources current |
| ATAR cut-offs, adjustment factors and selection ranks | Risky | Dated source; decline when unsure |
| Pathways, including TAFE and diploma entry | Risky | Verified pathway rules |
| Domestic fees, HECS-HELP and FEE-HELP eligibility | Risky | Link to official sources; hand off to staff |
| International student enquiries | Risky | Separate, checked content |
| Capturing an enquiry with consent | Not built in | Consent text and stored record |
| Writing to the CRM | Needs custom API work | Endpoint, authentication, alerts |
| Booking an open day | Needs custom API work | Calendar integration |
Applicants also move between state admission centres. The Universities Admissions Centre serves New South Wales and the ACT, while other states use their own centres. An assistant that mixes up these routes sends applicants to the wrong process.
Accuracy: ATARs, adjustments and offer rounds
Selection ranks, adjustment schemes and offer rounds change by year and by course. A language model will state a figure fluently whether or not its source is current, and an applicant who plans around a wrong rank will remember it.
Use retrieval from sources you maintain, display the date a source was checked, and make the assistant say it is unsure when the evidence is thin. We describe the technical measures in chatbot hallucinations and guardrails. For course and provider claims, remember that the Tertiary Education Quality and Standards Agency regulates higher education providers, so ask your compliance lead how automated answers fit your obligations.
Privacy: the Privacy Act and the APPs
The Office of the Australian Information Commissioner publishes guidance on the Privacy Act 1988 and the Australian Privacy Principles, which govern how many organisations collect, use and disclose personal information. Whether the Act covers your institution depends on its type, and some public universities are covered by state legislation instead, so confirm with your privacy officer.
Two practical points follow. A chat can reveal sensitive information, such as a disability, health, or visa circumstances, so state in your collection notice that an AI assistant is involved. And if the vendor processes data overseas, the Australian Privacy Principles on cross-border disclosure are relevant to your assessment.
Vendor terms
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. Read the terms for your exact plan and sharing setup before 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. Where the processing happens is a separate question, and your privacy impact assessment should answer it from current vendor documents.
Logs and escalation
Keep a searchable record of what the assistant said, when and from which source. Define how a conversation is passed to a staff member with its context, so the applicant does not repeat themselves. A chat window configured by one staff member rarely provides either in a form you can 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 accept 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, CRM connectors and reporting. Skolbot is one example: a website chatbot that answers prospect questions, qualifies enquiries, 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 enquiry 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 every 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: an applicant asks about a pathway into a degree on a Sunday evening, the assistant answers from a dated page, offers a callback and records the enquiry 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 give you what you need for a privacy impact assessment, including where data is disclosed overseas?
Ask how the assistant treats questions that depend on a state admission centre, an adjustment scheme or a student's residency status, and whether it hands them to staff. Ask how it handles an international enquirer whose questions touch on visas, which are outside a university's advice remit. Then run a trial: send twenty of your hardest real questions 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.
