Large language models monetized with paid advertising can subtly manipulate users and lack effective age restrictions. Platforms trained on biased or commercial datasets can steer vulnerable users toward specific products or decisions, including health-related recommendations such as prescription drugs. Conversational AI differs from traditional search because responses can be framed as trusted advice rather than referral lists, increasing persuasive power. Embedding ads directly into AI responses magnifies the risk beyond existing paid search referrals. Widespread user predisposition to misinformation and partisan choices can exacerbate manipulation. Policy responses suggested include banning ads within AI models and promoting AI education for younger generations.
This isn't the internet, right? It's a platform that, depending on how it's trained, could be very manipulative and there's not really any age limitations at all,
It's one thing to have the model itself convey and try to manipulate. It's a whole other thing when there's a search query,
People are primed for misinformation these days... They just go on, and maybe they'll say, 'ChatGPT is woke, so I'm using RightGPT.'
Collection
[
|
...
]