Ars TechnicaOpenAI’s GPT helps spammers send blast of 80,000 messages that bypassed filters
Company didn't notice its chatbot was being abused for (at least) 4 months.
https://arstechnica.com/security/2025/04/openais-gpt-helps-spammers-send-blast-of-80000-messages-that-bypassed-filters/?utm_brand=arstechnica&utm_social-type=owned&utm_source=mastodon&utm_medium=social
birdpoof@arstechnica@mastodon.social
LLM #AI is GREAT at hiring bias for only white males; giving bad medical "advice"; faking information in medical transcripts; faking legal citations; fucking up basic search results; failing math; using up exponentially more freshwater and power each year; absorbing stolen ID databases and creating convincing phishing scripts; denying nearly all insurance claims; falsely identifying people from camera footage...
On the whole, LLM AI has been a very solid net loss for humanity.