Skip to content

AI radio presenters: Who is behind the microphone?

Listen to Story · 5 min listen
Quick Takeaway · Key Summary

Between two songs, a short voice identifies the station, offers a few words and returns us to the music. It is one of radio’s most familiar moments. The voice no longer has to belong to someone speaking live in a studio: it might be prerecorded or generated synthetically.

The RadioTEDU AI English page, reviewed for this article on 8 November 2026, describes concise AI-hosted segments within its music schedule. The page also lists programmes such as TEDU Dawn, but does not specify how AI presentation is used in every edition. The station’s description raises a question for listeners: what decisions sit behind a short announcement?

Information comes before the voice

Sounding natural and being accurate are separate things. A station identification can be written and checked ahead of time. The next track’s name, a news item or an event time depends on information that may change. A fluent sentence can make a mistake harder to notice without making it true.

Imagine that a track-information feed is unavailable. A general, pre-approved transition may be more reliable than guessing the artist’s name. That requires a team to think beyond the voice generator: where did the information come from, and what happens when a dependency fails?

Once the words are ready, another set of decisions follows. Is a name pronounced correctly? Does the emphasis change the meaning? Can the announcement be understood over the music? Writing, producing speech and placing audio in a schedule are separate stages. If a real person’s voice sample is used, the team also needs to establish what recordings and distribution the person has agreed to.

How do listeners respond?

A small study from Türkiye offers one window into that question. Nihal Erdoğan Sepetci’s 2025 research draws on in-depth interviews with 12 people identified as listeners of an AI-presented Alem FM programme. The researcher reports positive views among those participants and motivations including curiosity, entertainment and experiencing something new.

📻
RadioTEDU Live Stream

Listen while exploring this story

Tune into campus radio with non-stop independent music and culture.

Because the interviews were conducted with people identified as programme listeners, they cannot explain why other people chose not to listen. Twelve interviews cannot establish the views of every radio listener, and the schedule described in the study should not be assumed current. The underlying question remains useful: is a synthetic voice interesting because it sounds natural, or because of the content and relationship the programme creates?

When an experiment sparked debate

Another story came from Poland in 2024. Associated Press reported that OFF Radio Kraków ended an experiment with AI presenter characters after about a week amid criticism following the dismissal of journalists. The station’s editor said the project had been intended to start a debate about AI.

That history does not predict the outcome of every synthetic-voice project. It does show why decisions about automation are connected to creative work and editorial responsibility. How a station explains those decisions can matter to its audience too.

Explaining how a voice was made

Presenting synthetic audio as though a person were speaking live can mislead listeners. A useful explanation can be more specific than “AI was used”: which segment contains a synthetic voice, who checked the words and which part was recorded?

Podcast platforms may also set their own conditions. Apple Podcasts’ guidelines require prominent disclosure of AI-generated audio or video in the content and metadata of each relevant episode and show. That is Apple’s rule; other services need to be checked separately.

RadioTEDU’s technology page presents its AI work alongside broadcast systems. For a campus station, the interesting question is not only how closely a voice resembles a presenter. It is what the tool does, which information needs checking and what happens when a component fails. Those human decisions remain behind the few seconds a listener hears between songs.

RadioTEDU monthly

The month in podcasts.

Receive podcast episodes from the latest 30 days and upcoming RadioTEDU ticketed events. No daily mail or unrelated announcements.

This consent covers only the monthly podcast newsletter and ticketed event updates. See our privacy notice.

Live RadioTEDU Pick a station and start listening
0:00 0:00

RadioTEDU · Big5 Türkiye Watch Party

RadioTEDU Live Screening: Big5 Türkiye Watch Party. Thursday 19 November, 18.00, D032. Click for free tickets.