Security

22 posts

Threat models, jailbreaks, guardrails, and safe AI adoption for teams.

RAG Poisoning and the Drift You Don’t See

Change the retrieval, change the answer. RAG turns inputs into control channels. Plant a sentence and the assistant will show it with confidence.

Security

Why prompt injecting through images works as a jailbreak, even for GPT 5, and how you can test on your own GPTs.

Jailbreak

It shouldn't be this easy to jailbreak GPT-5, but here we are with a new injection technique.

Jailbreak

Where you shouldn't be finding your private API key. The public internet.

Security

Review floods are a reputational DoS. How AI-made fakes, extortion, and pile-ons drown trust–and what founders can do to stay resilient.

Security

Studies show ~45% of AI-generated code carries OWASP-class flaws. The threat isn’t exotic - it’s ordinary defaults shipped at machine speed.

Security

AI scaled social engineering. Deepfakes, spoofs, and fake AI tools exploit meetings and calendars. Treat every channel as an attack surface.

Security

Indirect prompt injection can hijack agents and exfiltrate your docs. See how RAG, tools, and supply chains open leaks–and what to do next.

Security