Welcome to January 2025 Threat Highlight Report
The report reflects a surge in SaaS breaches by 300%, ongoing exploitation of network devices like Palo Alto firewalls, and active ransomware campaigns increasing by 139% compared to last year.
January report covers
The report highlights new vulnerabilities in AI frameworks such as Meta’s Llama, and introduces a novel jailbreak method for large language models to bypass security controls, increasing attack success rates by 60%. Additionally, the report notes security lapses in major cloud services like DeepSeek, exposing sensitive data, and emphasizes that cyber threats continue to evolve rapidly, targeting identity, AI systems, and network infrastructure.
The key findings in the January 2025 Threat Highlight Report include:
- SaaS breaches increased by 300% in 2024 compared to 2023.
- Ransomware attacks rose by 139% year-over-year, with groups like Cl0p and Lynx remaining active.
- Vulnerabilities were found in Meta’s Llama framework, notably to remote code execution via insecure deserialization.
- A new jailbreak method for large language models was discovered, increasing attack success rates by 60%.
- Security lapses in DeepSeek and Palo Alto firewalls exposed sensitive data and vulnerabilities.
- Threat actors continue exploiting network devices, cloud services, and AI systems with evolving techniques.
Be Ahead of the Game!
Stay informed about the latest cybersecurity threats and trends by subscribing to WithSecure's monthly threat highlights report!
Our comprehensive report provides an overview of last month's cybersecurity news, the changing threat landscape, and relevant advice.
Don't miss out on valuable insights - fill out the form to receive our report now!