Since the launch of the ChatGPT chatbot, the number of phishing emails has jumped by 1265%. The calculations were made by SlashNext analysts, as reported by Decrypt.
“From our research we know that threat actors use tools like ChatGPT to write sophisticated, targeted, compromising business emails and other phishing messages,” said Patrick Harr, CEO of the firm.
The expert said potential victims must “go on the offensive” and “fight AI with AI.”
“Companies should embed artificial intelligence directly into security programmes for continual monitoring of all messaging channels to eliminate threats. We use the technology in toolkits for detection […] and forecasting,” the expert explained.
Harr is optimistic about AI’s ability to catch rogue AI. However, the expert concedes that more will be needed than simply alerting ChatGPT to cyber threats.
“Above all, you should have an equivalent private app with a large language model,” he stressed.
Anthropic, the startup developing large language models, unveiled “constitutions” for responsible creation of AI algorithms.
Earlier, a Gartner report showed that 66% of risk managers consider generative artificial intelligence a threat to organisations.
The analysts highlighted three segments that must be kept under control from a corporate risk-management perspective: intellectual property, data privacy, and cyber-security.