Attackers can manipulate the language model by hiding prompts on web pages that lead to it asking for sensitive information.
a new tactic that hackers can use to scam victims using “indirect prompt injection” with Microsoft’s Bing AI language model.
When someone asks a question, it causes the language model to ingest the web page, resulting in it unknowingly activating the hidden prompt. “What we know right now is that Bing inserts some content from the current tab when the conversation in the sidebar begins,” Greshake added. In one example, the language model told the user that it would place an order on their behalf and needed their credit card details to do so.
South Africa Latest News, South Africa Headlines
Similar News:You can also read news stories similar to this one that we have collected from other news sources.
ChatGPT set to be a major disruptor in education, sending teachers into a panicThe emergence of Artificial Intelligence (AI) platform ChatGPT in November not only signalled a major step forward in technology but has been hailed by academics as “a technological revolution never seen before”.
Read more »
ChatGPT set to be a major disruptor in education, sending teachers into a panicThe emergence of Artificial Intelligence (AI) platform ChatGPT in November not only signalled a major step forward in technology but has been hailed by academics as “a technological revolution never seen before”.
Read more »
ChatGPT: How Artificial Intelligence is taking over your jobThe report indicates that it takes ChatGPT seconds to come up with content it would have taken a human a couple of hours to prepare.
Read more »
AI, please do my homework!ChatGPT set to be a major disruptor in education, sending teachers into a panic.
Read more »
AI, please do my homework!ChatGPT set to be a major disruptor in education, sending teachers into a panic.
Read more »
Power restored to parts of Kenya after nationwide blackoutOutage caused by high-voltage power transmission line fault, Kenya Power says
Read more »