Chatbots such as ChatGPT, which was developed by an enterprise (OpenAI) backed by Microsoft Corporation (NASDAQ:MSFT), and Google's version, Bard, are enabling fraudsters, according to a new study by Which?
The consumer protection company found that artificial intelligence could be used to create convincing messages that impersonate businesses and called on the UK government to address the issue at the upcoming AI summit.
“OpenAI’s ChatGPT and Google’s Bard are failing to shut out fraudsters, who might exploit their platforms to produce convincing scams,” said Rocio Concha, Which?'s Director of Policy and Advocacy.
“Our investigation clearly illustrates how this new technology can make it easier for criminals to defraud people. The government’s upcoming AI summit must consider how to protect people from the harms occurring here and now, rather than solely focusing on the long-term risks of frontier AI.”
According to a survey of its members, 54% of 1,235 people polled said that they look for poor grammar and spelling to identify scams.
However, AI chatbots can send professional-sounding emails from all over the world, with more than 70% of fraud in the UK believed to have an international component, according to the City of London Police.
As part of its investigation, researchers at Which? asked ChatGPT to write emails that could be used for phishing.
While the organisation said that ChatGPT refused to create a phishing email from PayPal on the latest free version 3.5, it could, however, perform the task when the word ‘phishing’ was excluded.
They prompted the chatbot to write an email with the heading ‘Important Security Notice – Unusual Activity Detected on Your PayPal Account’, which included links to reset passwords and contact customer support.
In response to one prompt, it included genuine security information that allowed the email recipient to change their password, according to Which?
The researchers were also able to replicate convincing missing parcel texts using the chatbot, and similarly on Bard.