As the world becomes more reliant on technology, cybersecurity has become a serious worry for individuals and organisations. There has been an increase in the usage of artificial intelligence to tackle cyber threats in recent years. Google is joining the trend by announcing its Cloud Security AI Workbench, a cybersecurity suite equipped with a specialized security LLM called Sec-PaLM.

At the RSA Conference 2023, Google unveiled its Cloud Security AI Workbench, which includes new AI-powered tools to enhance its customers’ cybersecurity. The suite has Sec-PaLM, a specialized AI language model “finely tuned for security use cases.” It is designed to aid in research related to software vulnerabilities, malware, threat indicators, and behavioural threat profiles.

The Cloud Security AI Workbench includes new tools such as Mandiant’s Threat Intelligence AI, which leverages Sec-PaLM to discover, summarize, and respond to security threats. Sec-PaLM also helps users of Chronicle, Google’s cloud cybersecurity service, to search for security events and interact “conservatively” with the results. Additionally, users of Google’s Security Command Center AI will receive a “human-readable” description of the attack exposure provided by Sec-PaLM.

VirusTotal Code Insight is the first tool available in the Cloud Security AI Workbench, and it is currently only available in a limited preview. Google has announced plans to roll out the remaining features to “trusted testers” in the coming months.

Google’s move into generative AI for cybersecurity follows in the footsteps of Microsoft, which announced its Security Copilot tool earlier this year. Security Copilot is a new tool aimed at “summarizing” and “making sense” of threat intelligence using OpenAI’s generative AI models, including GPT-4. Both Google and Microsoft claim that generative AI will better equip security professionals to combat emerging threats.

The use of AI in cybersecurity is an exciting development, and Google’s Cloud Security AI Workbench is a welcome addition to the market. However, the efficacy of generative AI for cybersecurity is yet to be fully explored. While it is a promising development, caution should be exercised until further research is conducted on its efficacy.