1️⃣ Breaking News

1. LockBit Ransomware Group Breached and Defaced

The notorious LockBit ransomware syndicate has reportedly been hacked, with its dark web site displaying a message mocking the group and linking to leaked internal communications. Cybersecurity experts suggest the breach is authentic and could significantly disrupt LockBit’s operations. The incident exposes the group’s aggressive tactics and may hinder its future activities. (Reuters)

2. UK Government Warns of Increased Cyber Threats Amid AI Adoption

At the CyberUK 2025 conference, UK Cabinet Office Minister Pat McFadden revealed that as artificial intelligence becomes more widespread, the country is expected to face an increase in both the frequency and severity of cyberattacks. In 2024, the National Cyber Security Centre received nearly 2,000 cyberattack reports, with 90 considered significant and 12 classified as highly severe—a threefold increase in major incidents from the previous year. The government plans to introduce a new cybersecurity strategy and legislate new powers under the upcoming Cyber Security and Resilience Bill. (Reuters)

3. Fake AI Tools Used to Spread Noodlophile Malware

Threat actors have been observed leveraging fake AI-powered tools as lures to entice users into downloading an information-stealing malware dubbed Noodlophile. These fake tools are promoted via legitimate-looking Facebook groups and viral social media campaigns, targeting users interested in AI-based video and image editing. The campaign has attracted over 62,000 views on a single post, indicating a significant reach. (The Hacker News)

2️⃣ Research Highlights

1. Open Challenges in Multi-Agent Security

A recent study introduces the field of “multi-agent security,” focusing on the unique threats posed by interacting AI agents. The research outlines potential risks such as secret collusion, coordinated attacks, and data poisoning, emphasizing the need for a unified approach to secure decentralized AI systems. (arXiv)

2. AI Predicts Chemical Transition States with High Precision

Researchers have developed an AI model capable of predicting chemical transition states with exceptional accuracy. This advancement could revolutionize computational chemistry by enabling more efficient drug discovery and materials design processes. (chemistryworld.com)

3️⃣ Featured Tools & Resources

1. OWASP GenAI LLM Top 10 Risks

The Open Worldwide Application Security Project (OWASP) has released a comprehensive guide detailing the top 10 security risks associated with large language models (LLMs). This resource provides insights into vulnerabilities such as prompt injection, insecure output handling, and sensitive information disclosure, offering mitigation strategies for each.

2. Exabeam’s Guide on LLM Security Risks and Best Practices

Exabeam has published a guide outlining the top security risks of using LLMs and seven best practices to mitigate these risks. The guide emphasizes the importance of runtime protection, monitoring, and implementing strict access controls to safeguard against information disclosure vulnerabilities in LLMs.

4️⃣ Bonus: Emerging Threats or Industry Events

Prompt Injection Attacks Highlighted as Top Security Threat

Prompt injection attacks have been identified as a critical security threat in the 2025 OWASP Top 10 for LLM Applications report. These attacks involve adversaries crafting inputs that appear legitimate but are designed to cause unintended behavior in machine learning models, particularly large language models (LLMs). The report emphasizes the need for robust safeguards to prevent such vulnerabilities.

—

Stay informed and vigilant as the fields of AI and cybersecurity continue to evolve rapidly.