Kevin Townsend

New physics-based research suggests large language models could predict when their own answers are about to go wrong — a potential game changer for...

Researchers demonstrate how multi-turn “storytelling” attacks bypass prompt-level filters, exposing systemic weaknesses in GPT-5’s defenses.

Splunk researchers developed a system to fingerprint post-logon behavior, using AI to find subtle signals of intrusion.

An AI extension to the Ox Security platform automatically generates organization specific code to fix vulnerabilities in the codebase.

As AI makes software development accessible to all, security teams face a new challenge: protecting applications built by non-developers at unprecedented speed and scale.

Cisco’s latest jailbreak method reveals just how easily sensitive data can be extracted from chatbots trained on proprietary or copyrighted content.

Should Gen Z to be treated as a separate attack surface within your company?

The global average cost of a breach fell to $4.44 million (the first decline in five years), but the average US cost rose to...

The need for secure encryption in IoT and IIoT devices is obvious, and potentially critical for OT and, by extension, much of the critical...

Experts unpack the risks of trusting agentic AI, arguing that fallibility, hype, and a lack of transparency demand caution—before automation outpaces our understanding.

Critics warn that a ban on ransomware payments may lead to dangerous unintended consequences, including forcing victims into secrecy or incentivizing attackers to shift...

With generative AI enabling fraud-as-a-service at scale, legacy defenses are crumbling. The next wave of cybercrime is faster, smarter, and terrifyingly synthetic.

Investigators from HMRC joined more than 100 Romanian police officers to arrest the 13 Romanian suspects in the counties of Ilfov, Giurgiu and Calarasi. 

The latest release of the xAI LLM, Grok-4, has already fallen to a sophisticated jailbreak.

‘Machine identities’, often used interchangeably with ‘non-human identities’ (NHIs), have been increasing rapidly since the start of digital transformation.

As organizations rush to adopt agentic AI, security leaders must confront the growing risk of invisible threats and new attack vectors.

The move could reshape how LLM developers gather information — and force new deals between creators and AI companies.

Rachel Tobac is a cyber social engineer. She is skilled at persuading people to do what she wants, rather than what they know they...

New research suggests more than 10,000 SaaS apps could remain vulnerable to a nOAuth variant despite the basic issue being disclosed in June 2023.

New "Echo Chamber" attack bypasses advanced LLM safeguards by subtly manipulating conversational context, proving highly effective across leading AI models.