The National Cyber Security has warned that chatbots such as ChatGPT could be used by cyber criminals to leak data and has said users shouldn't enter their private or sensitive information.
A UK intelligence agency has warned of potential security threats fromand other large language model-based chatbots, recommending users do not enter personal or sensitive information into the software.
Large language models are trained on huge datasets , but once that information is ingested, they do not continue to learn from the prompts entered by users – of which ChatGPT receives millions per day. As an example, the NCSC team writes: ‘A question might be sensitive because of data included in the query, or because who is asking the question . Examples of the latter might be if a CEO is discovered to have asked ‘how best to lay off an employee?’, or somebody asking revealing health or relationship questions.In addition to later use by the developer, queries could also be hacked, leaked or accidentally made public.
On the flaws in current generative artificial intelligence LLMs – as opposed to artificial general intelligence, the Holy Grail for developers – the NCSC highlighted a number of issues, including the bots ‘getting things wrong and ‘hallucinating’ incorrect facts’, showing bias and being gullible, and being ‘coaxed into creating toxic content’.
Deutschland Neuesten Nachrichten, Deutschland Schlagzeilen
Similar News:Sie können auch ähnliche Nachrichten wie diese lesen, die wir aus anderen Nachrichtenquellen gesammelt haben.
ChatGPT Chrome extensions are mainly junk - but I've found 4 worth installingWant to integrate ChatGPT into Chrome and make the most of the chatbot? We've got you 💪🤖
Weiterlesen »
Microsoft and GM deal means your next car might talk to youMicrosoft and GM deal means your next car might talk, lie, gaslight and manipulate you
Weiterlesen »
I think Amazfit's ChatGPT fitness watch is a terrible, dangerous ideaAmazfit's ChatGPT-powered fitness watch function is a disaster waiting to happen
Weiterlesen »
ChatGPT creator announces upgraded AI model that can 'see'ChatGPT proved revelatory when it released in 2022, threatening to upend everything from how students did homework to how software engineers wrote computer code. The software was based on a model called GPT-3.5, and now the company behind it has unveiled a new version.
Weiterlesen »
GPT-4 is bringing a massive upgrade to ChatGPTGPT-4 is here - here's everything you need to know
Weiterlesen »
OpenAI announces ChatGPT successor GPT-4The fourth version of the AI chatbot can process both images and text.
Weiterlesen »