Carnegie Mellon and Anthropic Explore LLMs in Cyberattacks
Jul 26
NVIDIA's GPUHammer Threat, Google AI's SQLite Save, & $13.5M for Zip Security - Cybersecurity AI Weekly #13
July 21, 2025 -
Cybersecurity AI Weekly
Hi there,
Welcome to this week's edition of Cybersecurity AI Weekly.
In this edition, we bring you a range of developments in the intersection of AI and cybersecurity. NVIDIA has issued a warning about a new threat called GPUHammer, which can disrupt AI model accuracy on their GPUs by causing memory bit flips. Meanwhile, Zip Security has secured $13.5 million in funding to bolster AI-driven cybersecurity solutions for businesses lacking dedicated security teams. Google has also made headlines with its AI agent 'Big Sleep' identifying a critical vulnerability in the SQLite database engine, potentially preventing significant exploitation.
Additionally, Empirical Security has raised $12 million to develop custom AI models for cybersecurity, and Resolver Inc. has introduced a new AI service to detect AI-generated child sexual abuse material. In collaboration with AWS, CrowdStrike is expanding its AI security tools on the AWS Marketplace, while LightOn has launched an AI platform for education in partnership with ENS Paris-Saclay. These stories highlight the ongoing efforts and innovations in using AI to enhance cybersecurity measures across various sectors.
Welcome to this week's edition of Cybersecurity AI Weekly.
In this edition, we bring you a range of developments in the intersection of AI and cybersecurity. NVIDIA has issued a warning about a new threat called GPUHammer, which can disrupt AI model accuracy on their GPUs by causing memory bit flips. Meanwhile, Zip Security has secured $13.5 million in funding to bolster AI-driven cybersecurity solutions for businesses lacking dedicated security teams. Google has also made headlines with its AI agent 'Big Sleep' identifying a critical vulnerability in the SQLite database engine, potentially preventing significant exploitation.
Additionally, Empirical Security has raised $12 million to develop custom AI models for cybersecurity, and Resolver Inc. has introduced a new AI service to detect AI-generated child sexual abuse material. In collaboration with AWS, CrowdStrike is expanding its AI security tools on the AWS Marketplace, while LightOn has launched an AI platform for education in partnership with ENS Paris-Saclay. These stories highlight the ongoing efforts and innovations in using AI to enhance cybersecurity measures across various sectors.
NVIDIA Alerts on GPUHammer Threat to AI Models
NVIDIA has issued a warning about GPUHammer, a new RowHammer attack variant that can significantly impact AI model accuracy on NVIDIA GPUs by causing memory bit flips. Read more
Zip Security Secures $13.5M to Boost AI Cybersecurity
Zip Security has raised $13.5 million in Series A funding to enhance cybersecurity for businesses without dedicated security teams, focusing on AI and automation. Read more
Google AI Identifies Critical SQLite Vulnerability
Google announced that its AI agent 'Big Sleep' discovered a critical vulnerability in the SQLite database engine, preventing potential exploitation. The vulnerability, CVE-2025-6965, affects versions prior to 3.50.2. Read more
Empirical Security Secures $12 Million for AI Cybersecurity Models
Empirical Security has raised $12 million in seed funding to develop custom AI models for cybersecurity. The funding, led by Costanoa Ventures, will help enhance threat detection and response by tailoring defenses to each organization's unique needs. Read more
Resolver Inc. Introduces AI Service for Detecting AI-Generated CSAM
On July 17, 2025, Resolver Inc. announced a new service using AI to identify and classify AI-generated child sexual abuse material, aiming to improve online child safety. Read more
CrowdStrike Expands AI Security Tools on AWS Marketplace
On July 16, 2025, CrowdStrike announced new AI security tools, falcon-mcp and AI Red Team Services, available on AWS Marketplace to enhance AI adoption in cybersecurity. Read more
LightOn and ENS Paris-Saclay Launch AI Platform for Education
LightOn has introduced its new generative AI platform, Paradigm Edu, in collaboration with ENS Paris-Saclay. This initiative focuses on enhancing administrative functions in higher education with a secure French cloud environment. Read more
StrongestLayer Secures $5.2M to Combat AI Email Threats
StrongestLayer has launched with $5.2 million in seed funding to address AI-driven email threats, focusing on enhancing email security. Read more
Abstract Security and SentinelOne Partner for AI-Driven Security
Abstract Security and SentinelOne have announced a partnership to integrate their technologies, aiming to improve threat detection and response by enhancing data processing and reducing costs. Read more
FINN Partners Launches AI Crisis Training Platform
FINN Partners has introduced 'CANARY FOR CRISIS', an AI-powered platform to aid communications teams in managing reputational threats and misinformation. Read more
Lansweeper Acquires Redjack for Asset Intelligence
Lansweeper has acquired Redjack, enhancing technology asset intelligence and visibility by combining their capabilities in network traffic analysis and asset discovery. Read more
Confident Security Secures $4.2M for AI Data Privacy
San Francisco-based startup Confident Security has launched its product, CONFSEC, which offers end-to-end encryption for AI interactions. The company has raised $4.2 million in seed funding to enhance data privacy in sectors like healthcare and finance. Read more
DebitMyData Unveils AI Security API Suite
DebitMyData has introduced a new LLM Security API Suite that leverages reinforcement learning and blockchain to improve AI security. The suite features tools like Agentic Logos™ and Agentic Avatars™ to address threats such as deepfakes and AI fraud, while ensuring regulatory compliance. Read more
Pinpoint Search Group and Spotlight.ai Partner for Sales Automation
On July 15, 2025, Pinpoint Search Group announced a partnership with Spotlight.ai to enhance sales execution for cybersecurity vendors by automating sales processes. Read more
Sysdig Exec Highlights AI's Impact on Cloud Security
At Sysdig Accelerate '25, Sysdig's Shantanu Gattani discussed how generative AI is heightening cloud security threats, emphasizing the need for rapid detection and response. Read more
Quest Software Introduces AI Tool for Active Directory Security
Quest Software has introduced Security Guardian Intelligence, a generative AI tool designed to enhance detection and response to identity threats in hybrid Active Directory and Microsoft Entra ID environments. The tool offers plain-language insights and guided remediation steps to improve response times. Read more
Cloud Security Alliance Introduces AI Controls Matrix
The Cloud Security Alliance has launched the AI Controls Matrix, a framework designed to mitigate risks in generative AI systems. It includes 243 controls across 18 domains to assist stakeholders in adhering to international AI regulations. Read more
We hope you enjoyed this article.
Consider subscribing to one of several newsletters we publish like Cybersecurity AI Weekly.
Also, consider following us on social media:
More from: Cybersecurity
More from: Vulnerabilities in AI
Subscribe to Cybersecurity AI Weekly
Weekly newsletter about AI in Cybersecurity.
