ChatGPT
Claude
Perplexity
Gemini
GrokAgentic AI acts on a goal; a generative chatbot replies
Agentic AI is software that works toward an objective and takes actions in your systems: booking a prospective student into an open day, updating a CRM record, following up on an incomplete application. A generative chatbot writes a reply and stops.
For a recruitment and admissions team, that shifts the risk. A wrong answer can be corrected in the next message. A wrong action (an open-day booking at the wrong campus, an overwritten record, a reminder sent to the wrong applicant) leaves a trace in your systems and with the student.
The label is also overused. Many tools marketed as "agentic" are chatbots under a new name. The quickest test is one question: does the tool write to your systems, or only read and answer? Our explainer on the definition of an AI admissions agent draws that line, and the guide to AI agents for student recruitment places it in the full applicant journey.
Three kinds of tool compared
Rule-based chatbots, generative chatbots and agents differ in what they can do after replying.
| Criterion | Rule-based chatbot | Generative chatbot | AI agent |
|---|---|---|---|
| Source of the answer | Hand-built decision tree | Language model, sometimes grounded in your site | Language model grounded in your content and data |
| Possible action | Send a link | Send a link, draft text | Book, record, follow up, hand off |
| CRM access | None | Often none | Read and write within a defined scope |
| Main risk | Question outside the script | Invented answer | Wrong action repeated at scale |
| Control needed | Script maintenance | Sources and guardrails | Action log, limited permissions, human escalation |
The last row changes how you operate. An agent needs access governance, not only a review of its wording.
Five enquiry workflows an agent can handle
Five tasks appear in almost every recruitment team, and an agent can carry them out without touching the offer decision.
Qualifying the first enquiry. A visitor asks about a degree, an intake or a campus. The agent asks the two or three missing questions (current qualification, intended start, domestic or international) and passes on a usable record rather than a bare email address.
Booking open days and campus tours. The agent offers a time, records the registration and sends the confirmation. This is where the gap with a chatbot that points to a form is clearest.
Following up on incomplete applications. A missing document, or an applicant who goes quiet after a first reply, is a silent loss. The agent sends the planned reminder rather than waiting for an adviser to check a list.
Keeping the CRM current. Course of interest, questions asked, stage reached: the agent writes them to the record so nobody re-keys them.
Handing over to a person. It passes the full conversation to an adviser when a request falls outside its scope. A poor handover cancels the gain from the other four, so test it before launch and track it with the KPIs for an admissions AI agent.
What stays with a person
The offer decision, appeals and complex personal circumstances stay human, for accountability as much as quality. That includes disability support requests and any question touching an applicant's visa or financial position.
For privacy, the Office of the Australian Information Commissioner has published guidance on commercially available AI products. Its guidance on privacy and the use of commercially available AI products says organisations should not enter personal information, and particularly sensitive information, into publicly available AI chatbots, and that businesses should make it clear when a public-facing chatbot is an AI rather than a human. It also says information collected through chatbots must still meet the Privacy Act's collection and notice requirements.
For an admissions team the lesson is practical. Do not paste applicant files into a public tool, give the agent only the data it needs, and write the collection notice before launch rather than after.
If you recruit in the European Union, EU rules can reach you. Annex III of the EU AI Act lists, as high-risk, systems intended to determine access or admission to educational institutions, and Article 14 requires that a person can override such a system's output. An agent that answers enquiries and books tours does not have that purpose; one that ranks or scores applications moves toward it. Ask counsel to classify it.
What this means in an Australian context
Australian admissions are shaped by the ATAR for school leavers, state-based tertiary admissions centres, and direct application routes for mature-age, postgraduate and international students. The Group of Eight universities, regional universities and private providers all recruit differently, and TEQSA regulates providers and the quality of their courses.
An agent must never suggest that it can see or change an application held with a tertiary admissions centre, and it should never imply a particular ATAR will secure an offer. Cut-offs and adjustment factors are published by the institutions and change from year to year.
Where it helps is in explaining entry pathways and prerequisites, answering international student questions across time zones, handling the surge of enquiries around ATAR release and offer rounds, and keeping up with the many enquiries that arrive after hours. For international recruitment, be careful with claims about visas and work rights: link to the official government pages rather than paraphrasing them.
Student finance is another trap. HECS-HELP and FEE-HELP eligibility depends on citizenship and course, so the agent should explain the general idea and point to the official source, never quote a personal outcome.
Five questions to ask before you choose a vendor
Ask every supplier these, including us.
- What can it write, and where? Request the list of possible actions and editable fields.
- Who can undo an action? A readable log and a rollback procedure are essential.
- Where is data stored, and does it train a model? The answer belongs in the contract.
- How is human escalation tested? Ask for a trial on real cases, not a scripted demo.
- How much of the claimed performance can be verified? Be wary of any result figure without a published method. No credible public benchmark currently sets a normal success rate for an admissions agent.
Where Skolbot fits
Skolbot deploys AI agents connected to the institution's CRM: the website agent answers from the institution's own content and passes the conversation and its qualification into the CRM. The CRM remains the source of truth, and the admissions team stays in control of the cases that need it. For the wider picture, read our guide to AI chatbots for student recruitment.
