← Back to glossary index

Recruitment Glossary

Chatbots

Recruitment chatbots are conversational interfaces that answer questions, collect information, guide applications, schedule activity, or support recruiters through text or voice. Some follow fixed rules, while others use natural-language or generative AI systems.

Recruiter Focus

The team should define exactly what the chatbot may say, collect, recommend, or change; disclose when a candidate is interacting with automation; provide an effective human route; and prevent the tool from inventing role terms, making unauthorised promises, or conducting hidden screening.

Why Chatbots Matters

A chatbot can give candidates useful information outside office hours and reduce repetitive administration. It can also create barriers for assistive technology or language differences, collect unnecessary personal data, misunderstand a request, expose confidential content, or make a candidate believe an automated response is an employment decision.

Terms Recruiters Commonly Compare

Live chat

Live chat connects the candidate with a human representative. A chatbot generates or selects responses automatically, even when the interface looks similar.

Recruitment automation

Recruitment automation covers any configured system action, such as reminders or stage updates. A chatbot is a conversational interface and may trigger automation, but the two concepts are not the same.

Recruitment Example

A careers-site chatbot answers approved questions about locations, application stages, benefits, and interview preparation. It can offer available interview slots but cannot alter eligibility, reject candidates, interpret health information, or negotiate pay. When confidence is low or an adjustment is requested, it creates a human-owned task and preserves the conversation under the approved retention rule.

When the conversation must move to a person

Escalation should be based on the effect of the request, not only whether the bot recognises the words. A technically fluent answer can still be inappropriate when the matter needs judgement, authority, or confidentiality.

  • A request for a reasonable adjustment or accessible alternative
  • A challenge to a decision, score, or use of personal information
  • A discrepancy involving identity, eligibility, safeguarding, or background screening
  • A proposed exception to approved pay, location, hours, process, or offer terms
  • Any answer the system cannot ground in a current authorised source

Implementation Playbook

  • Classify each intended function as information, administration, data collection, recommendation, or decision support and apply controls proportionate to its effect.
  • Use an approved knowledge source with version ownership, and test changed vacancies, closed roles, policy exceptions, and questions the bot should refuse.
  • Tell users they are speaking with a chatbot and make human contact available without requiring the person to disclose why.
  • Test keyboard, screen-reader, mobile, language, time-zone, identity, prompt-injection, and sensitive-data scenarios before release.
  • Monitor unresolved intents, escalations, inaccurate answers, complaints, and candidate abandonment; review transcripts only under defined access and retention rules.

Common Mistakes

  • Launching a chatbot because competitors have one without identifying a candidate problem it solves.
  • Letting generated answers draw from outdated or unapproved job content.
  • Using conversational questions to collect protected or sensitive information that the normal application would not request.
  • Offering a “speak to a person” path that is hidden, unstaffed, or returns the user to the chatbot.
  • Assuming a vendor’s general security statement covers the organisation’s prompts, integrations, candidate data, and retention settings.

Metrics to Track

Questions resolved with verified answers Human escalation completion time Inaccurate-answer incident rate Chat-to-application abandonment

Questions Recruiters Ask

Should a recruitment chatbot tell candidates it is automated?

Yes. Candidates should understand who or what they are interacting with, what the conversation can affect, how information is used, and how to reach an accountable person.

Can a chatbot screen candidates?

It may collect approved responses, but using them to rank or reject people creates a consequential selection system requiring much stronger validation, fairness, transparency, privacy, accessibility, and human-review controls.

What information should a recruitment chatbot avoid collecting?

Avoid information that is unnecessary for its defined function, especially sensitive personal details, confidential employer data, identity documents, medical explanations, or free text that no authorised process is ready to handle securely.

Sources and Review

ATZ CRM Recruitment Editorial Review · Reviewed 2026-08-05

Put Chatbots Into Practice with ATZ CRM

Use ATZ CRM to convert glossary concepts into daily recruiter workflows with sourcing pipelines, automation, scorecards, and reporting built for staffing and recruitment teams.