New rules being brought in for teens interacting with artificial intelligence still rely on people telling the truth, RTÉ's Work and Technology Correspondent Brian O’Donovan has told Behind the Story.
It comes as Meta - the parent company of Facebook, Instagram and WhatsApp - announced new safeguards for teens interacting with its AI tools.
Under the new controls, parents will be able to turn off their teens' access to one-on-one chats with AI characters and will also have more insights on how teens are interacting with AI.
Brian told Katie and David the companies want to be seen to move on this issue.
"The AI companies, the social media companies, [are] coming under some pressure to try to clean up their act when it comes to the safeguards and security and safety on their AI products," he explained.
Read more: Meta introduces new AI safeguards for teens
Brian pointed out that even the best safeguards could ultimately be bypassed.
"The biggest thing of all: have they admitted to being a teen at all in the first place?" he said.
"For you to be put into your Instagram teen account, you will have to have claimed that you are the age that you are – teenagers lie about these things all the time."
Meta has said it is aware of such issues and is "using AI technology to place those we suspect are teens into these protections, even if they tell us they’re adults".
Deepfakes
Brian said the problem with deepfake videos online, which are computer-created videos combined to create footage showing events that never actually happened, is growing.
"I’m getting these fake interviewers interviewing people in this kind of sketch show fake news stuff that’s meant to be of comedic value," he said.
"Some of it’s mildly funny – the quality has improved massively with these things.
"It’s a very short leap from getting a celebrity – living or dead – to say something funny to get a few clicks to taking a politician saying something they didn’t actually say, and it becomes this big problem.
"Will we reach a point soon where everybody just questions everything?"
Sexualisation of AI
It comes just days after Sam Altman, the head of ChatGPT, announced it will allow "erotica for verified adults" from December.
"His point [is] because we have changed the safeguards 'over here’, we can relax the safeguards ‘over here’ because we think we went too strict, people weren't liking it," Brian said.
"Now we will have more erotica [and] allow adults to behave like adults."
A recent survey by the US-based Centre for Democracy and Technology found one-in-five senior school students have or know someone who has had a romantic relationship with AI.
"It seems to be the sexualisation of AI to boost the user numbers, coupled with this utter confidence and belief by these tech companies that they have fixed the problems – so we don’t need to be censoring people anymore," Brian added.
You can listen to Behind the Story which is available on the RTÉ Radio Player.
You can also find episodes on Apple here, or on Spotify here.
Anyone affected by issues raised in this podcast can go to rte.ie/helplines
