Google's Sec-Gemini v1 Uses AI to Boost Cyber Defences

The cybersecurity threat landscape is growing in complexity as the increased use of remote work, cloud systems and open-source software has widened the attack surface significantly. As a result, organisations face threats from a broader range of actors than ever before.

These complications result in a defender-attacker asymmetry, where attacks only have to exploit one weakness and defenders must attempt to protect all systems simultaneously.

With this in mind, Google has announced the launch of its new experimental AI model that helps mitigate some of the strain security teams face from this imbalance  – Sec-Gemini v1.

This model will integrate near-real-time cybersecurity knowledge and tooling with Gemini’s advanced capabilities to pioneer the greater use of AI in cybersecurity.

In a LinkedIn post, Elie Bursztein, Cybersecurity Research Lead at Google, explains: “Very excited to announce Sec-Gemini v1, our experimental model specialised in cybersecurity, that will be made freely available to select organisations, institutions, professionals and NGOs for research purposes.

“Sec-Gemini v1 achieves state-of-the-art performance on key academic cybersecurity benchmarks, including threat intelligence knowledge (CTI-MCQ) and root cause mapping (CTI-RCM). 

I am looking forward to seeing it used to advance the AI cybersecurity frontier.”

Discover Sec-Gemini v1

Sec-Gemini v1 is effectively integrated with key data sources, OSV and Google Threat Intelligence.

As a result of this, it surpasses other models on key cybersecurity benchmarks and exceeds other models on CTI-MCQ by 11%.

The model works across critical workflows, such as threat detection, incident root cause analysis and vulnerability impact assessment. 

It can effectively answer cybersecurity questions due to its extensive integration with Mandiant Threat Intelligence data. 

Sec-Gemini v1 combines OSV and Mandiant data to offer an abundance of knowledge about several vulnerabilities and how attackers can exploit them. Organisations and analysts can utilise these outputs to dissect the risk and threat profile in their vulnerabilities.

Sec-Gemini v1 will be made available for professionals, institutions and NGOs for research purposes at no cost, highlighting Google’s commitment to collaboration across the cybersecurity industry.

What impact will it have on the wider cybersecurity industry?

Sec-Gemini v1 will set a benchmark for the use of AI in cybersecurity to tackle the asymmetry between defenders and attackers.

The model will improve the accuracy and efficiency of key SecOps tasks like vulnerability assessment and threat detection by combining near-real-time threat intelligence and Gemini’s advanced reasoning capabilities. 

Its ability to surpass industry benchmarks, such as CTI-MCQ and CTI-Root Cause Mapping, will secure its position as a vital asset for security executives.

Defenders can also respond with speed and ease as Sec-Gemini v1 offers contextual understanding of threats and vulnerabilities due to its integration with GTI, OSV and Mandiant. 

Sec-Gemini v1 will act as a catalyst for industry-wide innovation and resilience by encouraging collaboration across the cybersecurity ecosystem.

Explore the latest edition of Cyber Magazine and be part of the conversation at our global conference series, Tech & AI LIVE and Cyber LIVE.

Discover all our upcoming events and secure your tickets today. 

Cyber Magazine is a BizClik brand