The maker of ChatGPT is rolling out a range of new safety features today for under-18s who use the AI chatbot. For those with new teen accounts, OpenAI is removing the ability for the chatbot to respond using a human voice in order to minimise the tech's ability to appear like a person. It is also introducing regular reminders for young people to take a break from using the tool, reminding them ChatGPT is AI, and saying "it can wait".
The firm insisted this was not in response to a particular issue with children believing ChatGPT to be alive - but there are growing concerns about people becoming convinced that the AI tools they interact with are sentient. Last year Mustafa Suleyman, the head of AI at rival firm Microsoft, said "seemingly conscious" AI products "kept him awake at night" because he was so worried about their impact on society. In other areas, OpenAI seems to be loosely following the teen account model developed by the large social networks.
In a blog post announcing ChatGPT for Teens, external, the firm said teenagers' use of AI "should come with protections that reflect their developmental stage, reinforce real-world relationships, and support healthy use over time". It said teen accounts would become the default for those who identified themselves as teenagers when they signed up, and others whom the firm itself identified as under 18. But it did not elaborate on how it would identify these teens.
Under-13s are not allowed to use the product. Last year, ChatGPT launched Study Mode for students, in which it talks through schoolwork without giving away the answers. Young people will now be able to set Study Mode hours during which this will be the default setting.
They will also be able to activate "quiet times" when the tool will be switched off for them. OpenAI says of the young people who use ChatGPT, nine out of 10 use it to help them with learning. A report released yesterday, external by the Policy Exchange think tank claimed that the majority of UK universities are leaving themselves exposed to students enhancing their grades using AI by relying on online exams completed remotely.
It said 94% of the students it spoke to admitted to using the tech in assessments. OpenAI is also adding eating disorder-themed prompts to a list of harmful user requests which will be flagged to parents with linked accounts. The company said each alert would be reviewed by a human before it was sent, and it aimed to have them with parents within an hour of the request being made.
Extract — continue reading at the source.
