A new report published by WithSecure highlights another potential use of AI to create malicious content.
Researchers used GPT-3 (Generative Pre-trained Transformer 3) – language models that use machine learning to generate text – to create a variety of content deemed malicious.
The experiment involved phishing and spear phishing, harassment, social validation of scams, adopting a writing style, creating deliberate dissent, using mockups to create invitations for malicious text, and fake news.
“The fact that powerful, large-scale language models can now be accessed by anyone with an Internet connection has a very practical consequence: it’s now reasonable to assume that every new communication you receive might have been written with the help of a robot,” says Andy Patel, intelligence researcher at WithSecure who led the research. “The future use of AI to generate both harmful and useful content will require detection strategies that are able to understand the meaning and purpose of written content.”
The results lead the researchers to conclude that we will see the development of guest engineering as a discipline, as will the creation of malicious guests. Attackers are also likely to unpredictably expand the capabilities offered by large language models. This means that it becomes more difficult for platform providers to identify malicious or abusive content. Large language models already give criminals the ability to make any targeted communication more effective in the event of an attack.
“We started this research before ChatGPT made GPT-3 technology available to everyone,” adds Patel. “This development has increased our urgency and our efforts. Because to some extent we’re all blade runners now, trying to figure out if the intelligence we’re dealing with is ‘real’ or artificial.”
Source: WithSecure
And you?
What do you think ?
See also:
42% of marketers believe that AI will be the main trend in 2023, video streaming comes second followed by Metaverse and Connected TV
Fake mobile app posing as ChatGPT attempts to scam users about a $7.99* weekly subscription, App Store and Play Store reportedly contain several
Scientists say they are now actively trying to build sentient robots, but this brings new challenges and raises many ethical questions