Launching August 4 · Las Vegas
AI Resilience Center of Excellence
When AI fails — or is turned against you — resilience is what remains. A vendor-neutral center helping enterprises protect, monitor, contain, and recover in the age of AI agents. Founded with Rubrik, built with the industry.
CSAI Foundation AI Resilience Center of Excellence Founding Partner — contributing research from Rubrik Zero Labs, product telemetry, and executive engagement.
Enterprise AI is entering a new era of failure — from two directions
Data loss, destructive automation, and failed recovery will rise sharply as AI agents spread, and there are no fool-proof preventions in sight. Resilience is what lets an organization take the hit and keep operating.
The agentic surface is under attack
Attackers increasingly target AI agents, model-driven workflows, non-human identities, and automated enterprise systems — turning autonomy itself into an attack surface.
Systems fail on their own
Misalignment, excessive permissions, hallucinated reasoning, prompt injection, unsafe tool use, and rushed deployments produce failures with no attacker at all.
Nation-state actors using AI-enhanced insider threats, hardware-level vulnerabilities in AI compute, and the commoditization of frontier models through distillation all raise the urgency to build AI resilience now.
Built for the leaders accountable for resilience
Our first briefing
One paper anchors the launch — a problem statement to frame the discipline before we build it together.
Understanding AI Resilience in the Context of Organizational Resilience
Where AI resilience is genuinely new — and where it should build on decades of business-continuity, disaster-recovery, and operational-resilience thinking rather than reinvent it. A shared starting point for boards, CISOs, and risk leaders, co-authored with the Center's founding partners.
Published for August 4
Priorities we're building with the industry
This is a launch, not a finished catalog. These are the high-priority workstreams we're opening — shaped and authored with our Vanguard Members, Executive Advisory Committee, and the broader community.
Problem-statement briefing
The launch paper framing AI resilience within organizational resilience.
AI Resilience Reference Architecture
A shared model for designing recoverable AI-agent environments — a first working-group priority.
Board & executive briefings
Why AI resilience is a board issue, the questions to ask, and the metrics that matter.
Tabletop exercise kits
Simulations for data corruption, agent runaway, and recovery-failure scenarios.
Practitioner guidance series
Short, practical papers with maturity models and metrics for security and technology leaders.
Annual report & summit
A flagship annual report on the field's most important resilience issues, and a recurring summit.
Illustrative priorities for community input, shaped with our Vanguard Members and Executive Advisory Committee.
Practitioners set the agenda
The Center is guided by an Executive Advisory Committee of working CISOs, CIOs, Chief Data Officers, and risk leaders who ensure every deliverable reflects real operating reality — not vendor theory.
Join the Executive Advisory Committee →Part of a larger body of work
As a top-level program, the Center both feeds and draws on other CSAI initiatives.
Catastrophic Risk · Expert Group
The Center brings enterprise resilience and recovery expertise to the definition of catastrophic-AI controls.
Visit the project → Contributes toFrontier Ready Cybersecurity
The Center feeds recovery, containment, and continuity research into the initiative transforming cybersecurity for the frontier era.
Visit the initiative →Help found the resilience discipline
Join as a Vanguard Member, advise as a practitioner, or contribute to a workstream. The field is being built now — with you in the room.