Security | Threat Detection | Cyberattacks | DevSecOps | Compliance

In this first official episode of our Event Diary series, we take you inside AI Impact Summit 2026 at Bharat Mandapam, New Delhi. We had the chance to interact with a massive range of AI leaders—from visionary startup founders and engineers to data and compliance teams at major enterprises. The biggest takeaway? Companies are looking for ways to fast-track their compliance and enable their data safely. At Protecto, that is exactly what we’re solving.

Agentic AI is fundamentally changing how software operates, and in doing so, it breaks the identity and access models that many organizations still rely on. Unlike traditional applications, agentic systems are non-deterministic, long-running, and capable of autonomous decision-making across infrastructure, data, and production services. These systems do not fit within legacy identity assumptions built for humans, static workloads, perimeter controls, or long-lived credentials.

Your AI tools just became the perfect hiding spot for hackers. Cybercriminals have found a new attack vector: weaponizing the AI assistants your team uses every day. In this live demonstration, we show how adversaries can turn tools like Claude into persistent backdoors and how CrowdStrike Falcon stops them cold. TIMESTAMPS: WHAT YOU'LL LEARN:– How hackers exploit AI automation features to create backdoors– Why password resets and patches won't stop this attack– How behavioral detection catches threats hiding in legitimate tools– Real-time threat prevention in action.

Shadow AI models with weaker guardrails increase the power of already talented attackers and help less skilled users assemble dangerous campaigns. Frontier systems write malware when prompted as penetration tests, scale up attack speed and make it easier than ever to launch offensive operations against defenders. ⸻ For more information about us or if you have any questions you would like us to discuss email podcast@razorthorn.com. We give our clients a personalised, integrated approach to information security, driven by our belief in quality and discretion..

In this week's Intel Chat, Christopher Luft and Matt Bromiley discuss how attackers used AI to clone over 150 law firm websites, targeting fraud victims under the guise of offering legal assistance to recover lost funds. Chris points out how easy this has become with AI tools. Attackers can quickly clone a website, host it at a legitimate-looking domain, and start harvesting information. The episode also covers Russian cyber operations targeting the defense industrial base, Team PCP's campaign compromising 60,000+ servers, and exposed OLAMA AI infrastructure.