Key Takeaways:
- Microsoft Security Copilot shows 44% better accuracy for threat hunting and incident response
- Microsoft’s Cyber Signals report underscores the alarming trend of nation-state threat actors exploiting generative AI for malicious activities.
- Microsoft is actively exploring the deployment of AI-powered defenses to enhance cybersecurity measures and combat evolving threats.
Microsoft’s latest Cyber Signals 2024 report reveals the alarming trend of nation-state threat actors harnessing generative AI for malicious purposes. Microsoft has also outlined its proactive measures in deploying AI-powered defenses to tackle the ever-evolving threat landscape.
Microsoft Threat Intelligence team has reported that over 300 threat actors are currently using AI and Language Model Models (LLMs) in their attack operations. Cybercriminals are utilizing LLM technology to develop more sophisticated attack techniques such as automated attacks and deep fakes for phishing campaigns. Microsoft is closely monitoring the attack activity and has shared its findings with OpenAI to help detect and block AI-powered cyber operations.
Microsoft has outlined various methods it uses to safeguard customers from AI-assisted cyber threats. For example, it leverages behavior analysis and machine learning algorithms to identify and flag risky sign-ins and malware. Additionally, AI-based threat detection techniques are utilized to track any changes in resource usage and network traffic. Device health verification is performed to ensure safety before connecting to corporate networks.
“The use of LLMs in cybersecurity is a testament to AI’s potential. These models, when used for good, can analyze vast amounts of data to uncover patterns and trends in cyber threats, adding valuable context to threat intelligence. They assist in technical tasks such as reverse engineering and malware analysis, providing a new layer of defense against cyberattacks,” said Vasu Jakkal, Corporate Vice President of Security at Microsoft.
Microsoft Security Copilot shows 44% better accuracy for threat hunting and incident response
Microsoft has introduced several new security tools to help organizations enhance their defenses against sophisticated cyberattacks. According to customer feedback, Security Copilot has shown a 44 percent improvement in accuracy for various tasks such as threat hunting, incident response, intelligence gathering, and posture management.
Lastly, Microsoft has also announced several principles to mitigate the risk of state-backed threat actors using AI tools and APIs. These principles include identification and action against malicious threat actors, notifying other AI service providers, collaboration with stakeholders, and ensuring transparency. Microsoft says that these principles are based on their existing responsible AI practices and are part of their broader commitment to strengthening international laws and standards.