Google Are Putting a Focus on Cybersecurity and AI
Google’s new cybersecurity investment, Threat Intelligence, is bringing VirusTotal, Mandiant and Gemini together. As time goes on, people are trying to find more uses for generative AI, and that’s why Google is creating a new product which will utilize the Gemini 1.5 Pro language model. Google has openly said that this is going to reduce the time needed to reverse malware attacks. Gemini 1.5 Pro only took 34 seconds to break down the code of the Wannacry Virus. This is the ransomware that affected hospitals, organizations and companies across the world. This is incredible, although not surprising when you look at how adept AI is becoming.
Security Within the Online World
Even without this kind of technology, companies have been trying to keep their online customers safe for years. An example of this would be within the casino sector. When playing Lightning Roulette, you may notice that you need to verify your account when signing up and that you have to undergo a digital ID check. This cyber safety check is designed to ensure you can use the casino’s services, sampling titles like the above, in the smoothest way possible.
Even though the casino sector has been a leader in the sector of cybersecurity, Google is taking steps to improve online safety across different sectors as well. Gemini within the threat sector can be used so companies can find out how a potential cyberattack could affect them. This ensures that businesses are not under or overreacting to situations.
Users can also see the bigger picture when it comes to the landscape while prioritizing what it is they need to focus on. Mandiant on the other hand, is designed to help human experts who monitor malicious groups. VirusTotal is a community that posts threat indicators. When Google purchased Mandiant in 2022, they intended to use the experts from the team to assess vulnerabilities regarding AI.
Mandiant is used to test how secure AI models are. Even though AI is great at summarizing threats, while also reverse-engineering malware attacks, the models often become prey themselves. Data poisoning is a prime example here, which adds bad coding to the data that AI models are using, so they become unable to react to specific prompts.
Microsoft Have their Own AI Cybersecurity Software
Google is not the only company that is trying to combine AI with cybersecurity. Microsoft has since launched Copilot for Security. This is powered by GPT-4 and the company’s own AI model. Cybersecurity experts can ask questions regarding recent threats, and they can also connect with others. Whether or not generative AI has a good use within the cybersecurity sector is yet to be seen, but it seems like some of the biggest names in tech are making strides within the industry by trying to embrace it and the benefits it has to offer. Google’s new cybersecurity investment could help to cover all bases when it comes to security threats, providing an all-in-one program that not only identifies threats, but that could also take steps to eliminate them.
