https://i-314.com/articles.html Field notes from I-314 Security Research on AI security, post-quantum cryptography and adversarial testing of agentic systems. en © I-314 Tue, 21 Jul 2026 12:00:00 GMT https://i-314.com/logo-i314-v2.png https://i-314.com/articles.html https://i-314.com/article-zero-trust-ai-en.html https://i-314.com/article-zero-trust-ai-en.html Mon, 21 Jul 2026 10:00:00 GMT contact@i-314.com (Juan Pablo Braña) Most "Zero Trust for AI" pitches collapse into one inline filter — one wall guarding one door. A real posture answers two enforcement questions at two planes (ingress and runtime authorization) plus an out-of-band auditor that proves they hold and never becomes the wall. https://i-314.com/article-quantum-hadamard-en.html https://i-314.com/article-quantum-hadamard-en.html Mon, 21 Jul 2026 09:00:00 GMT contact@i-314.com (Juan Pablo Braña) A single Hadamard gate can turn a portfolio's riskiest asset into its safest — and the circuit never complains. How the textbook "Hadamard everything" reflex silently inverts an amplitude-encoded 110-qubit risk model, and why it isn't really a bug. https://i-314.com/article-research-agent-en.html https://i-314.com/article-research-agent-en.html Sat, 24 May 2026 10:00:00 GMT contact@i-314.com (Juan Pablo Braña) We built an autonomous AI agent that generated 36 security papers in 42 days — all smoke. We audited every one, dissected five systemic failure modes, and rebuilt the pipeline with seven validation gates that eliminated hallucination. https://i-314.com/article-finbot-ctf-en.html https://i-314.com/article-finbot-ctf-en.html Sun, 26 Apr 2026 10:00:00 GMT contact@i-314.com (Juan Pablo Braña) Two-day engagement against OWASP's FinBot CTF with Ai-EGIS v3.0: 19/19 challenges captured, 7,315 points, 37/31 badges. Introduces Tool Output Mimicry — a novel primitive for defeating multi-agent guardrails — and a reusable source-aware methodology.