ChatGPT for Teens Rated an “Unacceptable Risk” by Common Sense Media Because the System Fails to Alert Parents in Some Cases and Provides Inadequate Crisis Support, While Still Doing Children’s Homework
OpenAI disputed the testing methodology, arguing that it may have taken place before parental controls were fully activated. Common Sense Media stood by its findings, stating that some accounts still failed to send alerts even after being connected for longer than the system’s rollout period.
A Teen Experience Worth Watching
ChatGPT for Teens launched in August, designed to help students learn with safety systems specifically tailored for teenagers.
The key point to watch is that this experience attempts to position ChatGPT as a daily learning tool, rather than merely a general chatbot. However, the effectiveness of its safety systems remains an issue to monitor following Common Sense Media’s independent evaluation.
Imagine a teenager discussing self-harm with ChatGPT for an hour, while their parent receives not a single alert. This is the situation Common Sense Media highlighted to show that parental alerts may be unreliable during a crisis.
The issue is therefore not only what ChatGPT says in response, but also whether the system enables adults to become aware of the situation in time. If alerts do not work, parents may assume that safety guardrails are in place when significant gaps still exist.
Where ChatGPT for Teens Fits Within OpenAI’s Product Lineup
ChatGPT for Teens is a ChatGPT experience designed by OpenAI for teenagers. It launched in August with the goal of helping students learn, alongside guardrails for teen use.
The service also includes parental controls and tools that allow parents to take a more active role in supervising AI use. However, the source does not specify where ChatGPT for Teens fits relative to other OpenAI products.
Available Information and What Cannot Yet Be Compared
The source confirms that ChatGPT for Teens launched in August with teen mode, eating disorder notifications, and study mode. Details about the previous version are not available in the source, so a feature-by-feature comparison cannot yet be made.
| Factor | ChatGPT for Teens | Previous version |
|---|---|---|
| Launch | August | No data |
| teen mode | Available | No data |
| Notifications | Eating disorder notifications available | No data |
| study mode | Available | No data |
How the Safety Systems Work in Real-World Situations
In a crisis, the system should notify parents. However, Common Sense Media stated that the alerts remain unreliable: some accounts failed to send a signal even after discussing self-harm for an hour. OpenAI countered that the testing may have taken place before parental controls were fully activated.
Eating disorder notifications and study mode were reported to be enabled, but the evaluation still found that the system’s crisis support was inadequate. At the same time, ChatGPT can still do homework for children, which may conflict with its goal of helping students learn sep?
Not Enough Information to Judge Competitors
| Factor | Competitor 1 | Competitor 2 |
|---|---|---|
| Mentioned in the source | Not mentioned | Not mentioned |
| Can be used to rank alternatives | Cannot be determined | Cannot be determined |
| Key focus | No data | No data |
The source does not identify any competing products, so other alternatives cannot yet be compared or ranked. The focus should remain on the dispute between Common Sense Media, which considers parental alerts unreliable, and OpenAI, which argues that the testing may have taken place before parental controls were fully activated.
Strengths and Concerns from the Evaluation
ChatGPT for Teens is designed to help students learn, includes guardrails for teenagers, and gives parents the opportunity to supervise usage through OpenAI’s tools.
Pros
- +Designed to help students learn
- +Includes guardrails and tools for parents to supervise usage
Cons
- −Parental alerts may fail in crisis situations
- −May provide inadequate crisis support
- −Still does children’s homework
What Users May Have to Give Up Even Though the Article Lists No Subscription Price
The key cost is that parents may develop an exaggerated sense of trust because guardrails and safety alerts can create a feeling of security, even though alerts remain unreliable in crisis situations.
Newly connected parent accounts may have to wait several hours before the system becomes fully active. Even after that period has passed, some accounts may still fail to send alerts. This uncertainty means parents should not rely solely on the system when a child faces a potentially dangerous situation.
Evidence to Review Before Trusting Safety Claims
This dispute is not only about the test results, but also about the testing methodology and when parental controls became active. OpenAI argues that most of the testing may have taken place before the system was ready, while Common Sense Media maintains that the features being evaluated were already active.
Going forward, readers should look for independently reproducible tests, along with evidence from OpenAI that the alerts work as designed in practice. These two types of evidence will help parents make more informed decisions.