Bermuda Hundred Strategies

NEW: The AI Threat Modeling Field Guide

The definitive AI threat modeling field guide - from the original and leading global expert in AI threat modeling.

Mission-critical engineering, validated by mathematical proof.

A critical guide for all AI practitioners at all levels of the organisation.

Angles of Attack: The AI Security Intelligence Brief

The AI & Quantum Security Intelligence Brief is published by Shoshana on Substack at disesdi.substack.com and is recognised as a leading voice in the field of AI security.

Subscriptions are currently $7 per month, $77 per year.

Pricing will rise on 10th September 2026.

Subscribe →

General Consulting

Shoshana consults on all aspects of AI deployment from strategy to security implementation, including rapid analysis of how and where AI should be deployed in your organisation to maximise the benefits, and minimise the risks.

To discuss consulting projects and any other custom requirements, please email us at contact@bermudahundred.io, or book a no-obligation exploratory call to discuss whether and how Bermuda Hundred can fulfil your needs.

Book a call →

About Shoshana Cox

Shoshana Cox is an AI security architect, researcher, and strategist with over a decade of experience in mission-critical AI. Her professional background spans roles as a mathematician, red team lead, data scientist, and Chief Data Officer.

Currently serving as the CEO and Head of Research at Bermuda Hundred Strategies, Shoshana specializes in the intersection of AI, national security, and threat modeling. Her work focuses on bridging the gap between probabilistic systems and deterministic security requirements, providing high-level strategy for C-suite executives, and threat modeling training for technical teams.

Shoshana is widely recognized for her work on AI red teaming, MLSecOps and AI defensive architectures, authoring technical papers and holding a patent in AI security (US 12,093,400 B1). Her 2022 & 2025 papers on adversarial AI attack & defense were validated by a NIST-published mathematical proof & resulting recommendations, representing a rare post-hoc mathematical confirmation of engineering work.

Shoshana is a member of the core author team for the OWASP AI Exchange, where she contributes to international engineering standards, technical requirements for the EU AI Act, and advises on national security issues relating to AI. She is an active voice in the industry through her Substack newsletter Angles of Attack, where she provides in-depth analysis on topics like agentic threat modeling, AI red teaming, and the security risks of AI.

Some highlights of Shoshana’s published works:

“Securing AI/ML Systems in the Age of Information Warfare”, field’s first MLSecOps paper in 2022 prior to ChatGPT

“Quantifying The Risk Of Transferred Black Box Attacks” in 2025, mathematically demonstrating AI red teaming was invalid for security

US Patent US12093400B1, “Systems and methods for model security in distributed model training applications”

Shoshana’s core work on AI security was validated by NIST (The National Institute of Standards and Technology) mathematical proof in 2026.

Media/Speaking

Shoshana has spoken at numerous conferences in the AI, security, and intelligence fields.

She is selectively available for podcasts, television, and other media work.

For all enquires about Shoshana’s availability please email us at contact@bermudahundred.io or book a call.

Book a call →

Contact

For consulting, partnership, or other business enquiries, email us at contact@bermudahundred.io or book a call.