When in November OpenAI launched the revolutionary artificial intelligence language model ChatGPT millions of users have been fascinated by its capabilities. Generative Pre-trained Transformer (GPT) models brought generative AI to the fore, reaching 1 million users in five days and OpenAI ranks among the top 150 most visited sites and apps in Italy. However, while OpenAI is committed to ensuring that its technology is used appropriately, the capabilities of theartificial intelligence have also increased negatively becoming one instrument used for to ease the management of cyber attacks and cybercrime.
ChatGPT exploited by cybercrime
CybergON, the Elmec Informatica business unit that deals with cybersecurity has highlighted the main ones attack techniques conducted through AI and models Generative Pre-trained Transformer (GPT) that offer more speed e automation to cybercriminals and, at the same time, greater accessibility even to people without technical skills.
Use ChatGPT Confidently With NordVPN – Up To 63% Off For 2 Years Of NordVPN
According to the evidence found by CybergON, artificial intelligence – and more specifically ChatGPT – can be used to 3 main malicious purposes: to improve social engineering techniques and make the more targeted phishing attacksper install malware and to create some DeepFake campaign.
How does it work
Starting from its main function, which is that of a chatbot, adding an advanced text interface and the recording of images, video and audio, the technology can create a video output that represents the person reciting sentences that he never actually said. In this case, DeepFakes can be used in defamatory campaigns, fraud, theft and wholly illegal impersonation.
However, while there are risks, artificial intelligence can be used also to advance cyber security. For example, vendors can locate and fix vulnerabilities faster , in order to make the use of software safer and avoid leaving doors open to potential attackers. Furthermore, it can be integrated into the anti-phishing training of a company to raise employee awareness.
ChatGPT and cybercrime: the opinion of Elisa Ballerio, Marketing Director of CybergON
“The attack surface in modern enterprise environments is ever increasing, as are the volumes of data that need to be monitored and analyzed for potential risks,” he says. Elisa Ballerio, Marketing Director of CybergON. “A new level of solutions is needed to counter cyber threats: AI is the tool that, when combined with the human factor and other security technologies, can provide the greatest protection. At the moment, AI alone is not enough, for this reason, to verify the degree of the threat, as well as to determine the social and economic impact, human experience must come into play”.