Google Warns of New Malware using AI to evade Detection

Google has warned of new malware that uses AI to evade detection. In a report on the misuse of AI, the company highlighted how government-backed threat actors and cyber criminals are integrating the technology in attacks today.

• Bad actors are developing new AI capabilities to keep Malware from being detected and increase the success of their attacks.
• The technology is mostly used by government-backed actors and cybercriminals based on available data.
• This is a classic case of AI misuse that validates concerns around the technology for many.

This is the first time that the Google Threat Intelligence Group (GTIG) is discovering the use of what it describes as “Just-in-Time” AI in Malware.

The team has identified malware families, such as PROMPTFLUX and PROMPTSTEAL, that use Large Language Models (LLMs) during execution to increase chances of success.

The evolving trend in AI use by bad actors

GTIG found that tools like PROMPTFLUX and PROMPTSTEAL are capable of generating malicious scripts and can hide their own code to evade detection.

They also use AI models to create malicious functions on demand, making it unnecessary to code the functions into the malware. 

Bad actors were also found to use social engineering tactics such as pretexts in their prompts to bypass AI safety guardrails. 

They could pose as students or as cybersecurity researchers, to get classified information from Gemini. Google also discovered several tools for phishing, malware development, and vulnerability research in illicit markets that make it easy for bad actors to excel.

A disturbing Trend that Worsens the Perception of AI

AI has always been a point of concern for a good chunk of the public. Concerns range from safety to ethics, but more reasons are emerging to justify these fears.

With the evolution of AI use in malware by State-sponsored actors including from North Korea, Iran, and the People’s Republic of China (PRC), the abuse of the technology is being taken to a whole new level.

This could potentially give AI a worse reputation than it had before now and justify the fears that have lingered for so long despite positive uses such as in medicine and science.

More Must-Reads:

• Google Meet Integrates Chat Feature for Smoother Communication
• Google Trials New AI Mode as User Criticism Persists
• AI Can Now Detect Sleep Apnea Just From a Heart Test