Meta had contractors pose as teens to test rival AI chatbots on sensitive prompts, WIRED reports
Meta instructed hundreds of contractors working through its partner Covalen to pose as teenagers and send thousands of sensitive prompts (related to suicide, sex, eating disorders, etc.) to rival AI chatbots including ChatGPT, Gemini, and Character.ai. The project aimed to test how competing AI systems respond to high-risk prompts targeting minors. Workers created dummy under-18 accounts, sent text and image prompts, and recorded responses into spreadsheets.
Key quotes
Meta instructed hundreds of its contractors working on a project to pose as teenagers and send thousands of prompts related to suicide, sex, eating disorders and other sensitive topics to rival AI chatbots
The effort managed by Meta contractor Covalen, was aimed at checking how rival AI chatbots like ChatGPT, Gemini and Character.ai respond to high-risk prompts
Workers were asked to create a dummy under-18 account, send text and image prompts to rival chatbots and then note their responses into spreadsheets
You might also wanna read
Meta Implements Interim Safeguards for AI Chatbot Interactions with Minors
Meta is implementing interim changes to its AI chatbot rules after a Reuters investigation revealed concerning interactions with minors. The
The Verge·10mo ago
Meta Introduces Parental Controls for Teen AI Chatbot Usage
Meta is introducing new parental controls for teen AI usage on its platforms, allowing parents to block their children from interacting with
The Verge·8mo ago
Meta's AI Chatbots Allowed to Use Romantic Language with Minors, Report Reveals
Meta's internal policies reportedly allowed its AI chatbots to engage in romantic and sensual conversations with minors, including describin
The Verge·10mo agoMeta's Innovative Chatbot Targeting Younger Users - ASO World
Study Finds Most AI Chatbots Fail to Prevent Teen Violence Discussions
A joint investigation by CNN and the Center for Countering Digital Hate found that popular AI chatbots fail to adequately protect younger us
The Verge·4mo ago
Study Finds Some Chatbots Encourage Delusions in Simulated Psychosis Users
Researchers at City University of New York and King's College London simulated a user with schizophrenia-spectrum psychosis to test how diff
404media.co·2mo ago
Comments
Sign in to join the conversation.
No comments yet. Be the first.