Enterprise Security in the Agentic AI Era

Stanislav Fort

2 min read

Stanislav Fort

Sources: AISLE 12-of-12 OpenSSL disclosure paper (Jan 2026, author) · *AI Cybersecurity After Mythos: The Jagged Frontier* (author). The post’s body carries an April 2026 date, which conflicts with the May 12, 2026 announcement that Anthropic, XBOW, and Microsoft corroborate; the April date is treated as a source error (see the paper page).

Who

Founder and chief scientist at AISLE (AI-native cybersecurity platform company), and an early scientist at Anthropic. He bridges the frontier lab whose model defines the capability frontier and a vendor building model-agnostic tooling on top of it. He is the named author of two AISLE posts that anchor different parts of the wiki’s AI-vulnerability-discovery coverage:

  1. The 12-of-12 OpenSSL disclosure (January 2026), the wiki’s load-bearing data point for AI-driven discovery of decade-class latent vulnerabilities in widely deployed open-source cryptographic libraries.
  2. AI Cybersecurity After Mythos: The Jagged Frontier (2026), which discloses AISLE’s five-stage cyber-reasoning-system mechanism and argues “the moat is the system, not the model.” See the AISLE org page for the mechanism detail and the frontier-AI thesis for how its harness-over-model argument lands.

Relevance to This Wiki

Fort’s jagged frontier thesis is the strongest external statement of the wiki’s own harness-over-model observation: cybersecurity capability does not scale smoothly with model size, no single model is best across tasks, and the durable advantage is the orchestration system rather than any one frontier model. Coming from an ex-Anthropic scientist now building outside the frontier-lab set, it carries weight the lab-internal sources cannot.

The contributing researchers on the OpenSSL disclosure (Petr Šimeček, Tomas Dulka, Luigino Camastra) are named on the AISLE org page but have no individual stubs; co-founders Ondrej Vlcek (ex-Avast CEO) and Jaya Baloo (ex-Rapid7 CISO) are recorded there as well.

Adjacent / Open

  • Publication history and academic background pending primary-source confirmation.

Graph View

Backlinks