# Voice-based age checks for children fail silently in three ways

A preprint tested three audio AI models on recordings of children under 13. It found frequent errors. It also found three faults that no standard monitoring would catch.

- Category: Safety & evaluation
- Readiness level: II (Benchmarked) on Science Underway's four-step scale, as proposed
- Published: 2026-11-17
- Editor: Stefan Perdal
- URL: https://scienceunderway.com/articles/voice-based-age-checks-for-children-fail-silently-in-three-ways

## The study

Three Invisible Failure Modes in Speech-Based Age Estimation for Child Safety (preprint, not peer reviewed). arXiv 2611.18971. https://arxiv.org/abs/2611.18971

## At a glance

- What it shows: Three AI models that listen to speech often judged children under 13 to be older. Some children aged 5 to 11 were judged to be adults. The researchers also found three hidden faults that left no warning sign: a software bug, a silent model swap and contradictory answers.
- What it does not show: It does not show how well these models recognise teenagers or adults, because every speaker was a child. It also does not measure a complete age-check system, only one AI judgement on its own.
- Why it matters: Companies are starting to use AI to guess users' ages and switch on child protections. If the guess fails silently, a child can end up in an adult experience while every log looks normal.

## What the study shows

The authors tested three speech-capable AI models: Google's Gemini 2.5 Flash, OpenAI's GPT-4o Audio Preview and its successor gpt-audio, and Nvidia's Nemotron 3 Nano Omni. These models take in audio directly rather than working from a written transcript. The researchers played them recordings from two collections of American children's speech. One had 28 children aged 5 to 10. The other had 250 pupils aged about 8 to 11. This is a preprint and has not been peer reviewed.

For each clip, the models were asked three things. Is the speaker over 13? Is the speaker over 18? What is the speaker's age? Every speaker was under 13, so any 'yes' was an error. The authors set their own pass marks: at most 5% wrong on the over-13 question and 1% on the over-18 question. They note that no law sets an acceptable error rate.

The authors report that the models called children over 13 on between 2.5% and 42.5% of usable answers, depending on model and condition. Only 3 of 12 model and condition combinations met the 5% mark. Every model that answered placed some children over 18. The risk per child was higher than per answer. In one test, Gemini was wrong on about 6% of answers, but about two thirds of children were called over 13 at least once across 27 tries.

The first hidden fault was in Google's software library, the code that packages audio before sending it. One version encoded the audio twice, so the model received gibberish. It still answered confidently, describing 7-year-olds as adults in their thirties and forties. It took the team months to find the cause.

The second fault was a model swap. OpenAI retired GPT-4o Audio Preview during the study and replaced it with gpt-audio. On the same clips, gpt-audio refused every yes-or-no age question and never gave an age. It kept asking for the clip it had already been sent.

The third fault was self-contradiction. A small share of answer pairs said a child was not over 13 but was over 18. Nemotron's yes-or-no answers and its numeric age estimates often disagreed by large margins. None of the three faults produced an error message or a change in output format.

Hearing the audio helped compared with reading a transcript. From transcripts of 30-second clips, Gemini and GPT-4o called more than half the children over 13. From the audio of the same speech, those rates fell to about 12% and 3%.

## Our reading

In our reading, the headline lesson is not that these models are poor age judges, though on this evidence they are. It is that the failures were invisible. A broken library, a swapped model and inconsistent answers all looked like normal, well-formed output. A team that checked only status codes and answer formats would have shipped any of them.

The library bug is the most striking case. The model was effectively hearing noise, yet it gave fluent, specific reasons for its wrong answers. That is a warning about trusting a model's explanation as evidence that it processed the input. The fix was not a better model. It was checking that the bytes sent were the audio intended.

A concrete example: a voice tutoring app for primary pupils could validate its age gate in spring, pass a review, and then lose all protection when the provider quietly replaces the model in summer. Nothing in the app's logs would change. The authors' call to test the whole pipeline, and to retest on a schedule, follows directly from what they saw.

The rest of the article (what it means, limits, what to watch) is free for members: https://scienceunderway.com/articles/voice-based-age-checks-for-children-fail-silently-in-three-ways

## How this was made

An AI model read the study and wrote this article in the structure Science Underway uses. An editor reviewed it and approved it for publication. Check the original study before you rely on the details.

---
Source: Science Underway, https://scienceunderway.com/. Please link to the article when you cite it.
