
Spammers Forgot to Delete the Chatbot's Apology, and Now the Whole Internet Knows
Ourtake
The spam networks left the chatbot's apology in the posts, which is a handy tell for anyone looking for AI-written junk.
What happened
Fake Amazon product reviews and a Twitter spam network of at least 59,645 accounts were found posting ChatGPT-generated text, often including the chatbot's error phrases such as 'as an AI language model'. Twitter suspended the accounts Motherboard examined, and Amazon said it bans and takes legal action against fake review brokers.