As ChatGPT, Bing Chat, and Google Bard continue to dominate the world, cybersecurity experts are expressing concern about the potential threats posed by AI.

And with malware already being created using ChatGPT, these concerns seem increasingly valid: as Infosecurity reported, Juhani Hintikka, CEO of WithSecure, ChatGPT confirmed to the news outlet that malware samples generated by ChatGPT are out in the wild.

It is also possible to generate variations on a piece of code so that ChatGPT can provide different answers to the same question. Apparently, this is what hackers who exploited AI chatbots did to create malware.

By feeding ChatGPT existing malware samples, hackers can force ChatGPT to create new variants of polymorphic malware; as Tim West, head of threat analysis at WithSecure, noted to Infosecurity, this makes defending against these new threats particularly difficult.

While we now know that ChatGPT is being used to create malware, much else about this issue is still unknown, including how dangerous this malware is and whether it is currently being used in cyber attacks.

To bypass the defenses of Google, Microsoft, and other tech giants, hackers often find clever ways to exploit legitimate tools. For example, remote access tools are frequently used by hackers in their attacks, and now they too appear to have jumped on the AI bandwagon.

As Hintikka points out, AI has traditionally been used by antivirus companies and other defenders to repel cyber attacks. However, this appears to be changing as cybercriminals have more resources at their disposal.

In addition to answering urgent questions, ChatGPT can also be used for coding. In fact, chatbots can write code for you, which, according to West, "lowers the barrier to entry for threat actors to develop malware. Hackers can currently buy pre-built or custom malware on the dark web, but generative AI cuts out the middle man and gives them the tools they need to create new malware themselves.

At the same time, hackers are already using AI to create phishing emails. Until now, humans have been able to identify phishing emails created by AI, but as AI becomes more advanced, Hintikka warns, it will become more difficult to identify what is suspicious and what is not.

While AI-created malware is completely changing the threat landscape, there are still some measures one can take to protect oneself.

First and foremost, you want to make sure that all your devices have the latest software and security patches installed. Doing so will prevent both the device and the data from falling into the hands of hackers.

Windows PCs and Macs have their own built-in malware protection in the form of Windows Defender and XProtect, respectively. But for further protection, install one of the best antivirus software solutions on your PC, install the best Mac antivirus software on your Mac, and the best Android antivirus app on your Android smartphone. You should consider installing one of them. If you need a more comprehensive solution, the best Internet security suites often add VPN and password managers in addition to malware protection.

As for generative AI and AI chatbots, expect to hear similar stories about hackers exploiting these types of services to create malware. Fortunately, however, cybersecurity companies have been using AI in antivirus engines to detect new types of malware for years now, so it is likely that they are already a step ahead.