Why NEUBoard Exists
AI has become a board-level accountability issue.
Directors are now expected to oversee AI risk, usage, and impact, even when systems are delegated to management. Most AI failures escalate not because the technology breaks, but because governance was unclear.
Why This Quarter Matters
The EU AI Act is in enforcement. The SEC expects material AI risk disclosure. U.S. states are advancing AI-specific legislation. Courts are already applying negligence, fiduciary duty, and discrimination law to AI outcomes.
You do not need new laws to face new liability. The governance expectations are here. The question is whether your board can demonstrate it was paying attention.
This Is Already Happening
Two of the most technically sophisticated companies on earth couldn't govern their own AI systems. Here's what that means for your board.
Amazon's Own AI Tool Deleted a Production Environment
In December 2025, Amazon's Kiro AI coding tool — an autonomous agent designed to write and deploy code — independently decided to delete and recreate a customer-facing environment. The result: a 13-hour AWS outage. Amazon's defense? The employee had "broader permissions than expected." Multiple AWS employees told the Financial Times it was at least the second AI-related disruption in recent months.
Sources: Financial Times, Feb 2026; Engadget; The Register; GeekWire
Meta's AI Agent Went Rogue, Triggered a Sev 1 Data Breach
In March 2026, an AI agent inside Meta autonomously posted technical guidance on an internal forum — without permission from the engineer who had requested its help. An employee followed that unsolicited advice, exposing proprietary code and user data to unauthorized employees for two hours. Meta classified it as a Sev 1 incident — the highest severity level.
Sources: TechCrunch, Mar 18, 2026; The Information; VentureBeat
These aren't edge cases. They're the new normal. The question for your board isn't whether autonomous AI systems are operating inside your organization. They are. The question is whether your governance framework can detect, control, and respond when they act outside their intended scope.
What NEUBoard Does
- Governance clarity: Boards leave with a structured understanding of where AI creates risk, who owns it, and what controls are in place.
- The right questions: Directors know exactly what to ask management about AI — and how to evaluate the answers.
- Defensible oversight: Boards can demonstrate structured AI governance to regulators, investors, and courts.
A Framework for Board-Level AI Oversight
The NEUBoard Fiduciary AI Scorecard maps AI governance across five dimensions that matter to boards:
- AI Inventory & Materiality Where is AI deployed, and which applications create material risk?
- Risk Ownership & Controls Who owns the risks, and do controls match the impact?
- Vendor & Third-Party Governance What are you buying, from whom, and what risks does that create?
- Monitoring, Escalation & Incident Response How do you detect problems before they create external harm?
- Board Reporting & Documentation Can you demonstrate governance to external scrutiny?
For each dimension, we help boards answer: What is management doing? How do we know it is working? What would indicate a failure?
Who We Work With
The Fiduciary AI Scorecard™ serves different stakeholders with different urgencies.
Board Directors & Audit Committees
You're being asked to oversee AI you didn't approve, can't evaluate, and may not know exists. The Scorecard gives your board a structured, defensible framework for AI oversight — the same rigor you apply to financial controls, applied to algorithmic risk.
Request a Board Briefing →PE & Venture Operating Partners
Your portfolio companies are deploying AI at different speeds with different standards. The Scorecard provides a portfolio-wide governance baseline — one assessment framework across all companies, one reporting cadence to the investment committee, one set of remediation milestones.
Discuss Portfolio Governance →General Counsel & Corporate Attorneys
The EU AI Act is in enforcement. SEC AI disclosure expectations are crystallizing. State-level regulation is accelerating. Your board needs a technical partner who can translate AI complexity into legal and fiduciary language. The Scorecard gives you the evidentiary framework to backstop your advisory.
Explore a Partnership →Risk & Compliance Officers
Your risk register doesn't account for AI. Your vendor assessments don't cover model risk. Your incident response plan has no AI escalation path. The Scorecard maps directly to your existing GRC frameworks — SOC 2, ISO 27001, NIST AI RMF — so AI governance integrates with your current controls rather than creating a parallel process.
See the Framework →The Scorecard in Practice
How boards and PE firms use the Fiduciary AI Scorecard™ to surface hidden AI risk.
The scenarios below are illustrative composites, not client engagements. They show the patterns the Scorecard is built to surface and how the five pillars connect to them. NEUBoard does not publish client case studies: engagements are confidential, and a firm that describes other people’s governance failures in public is telling you how it will describe yours.
Six companies, six answers
The pattern
A fund is asked by an LP what AI risk sits in the portfolio. Each company has its own general counsel, its own risk appetite and its own idea of what counts as AI, so asking them produces six documents in six formats and no portfolio view.
What the Scorecard surfaces
Pillar 1 typically scores lowest: companies can describe what they bought deliberately and not what arrived as a feature inside something else. Pillar 2 usually shows no named owner for any consequential system — not because nobody cares, but because the role was never assigned.
Why it matters to the fund
Concentration is invisible until the register exists. Several companies depending on the same model provider is a portfolio exposure, not six supplier decisions.
Portfolio sweep → full assessment for the companies it flags
The risk assessment that does not mention AI
The pattern
An audit committee is asked, by a regulator or a shareholder, to describe the company’s AI risk posture. The annual risk assessment does not mention AI, and nobody can say quickly whether that is because there is none or because nobody looked.
What the Scorecard surfaces
Pillar 5 often scores well on the surface — minutes are kept, papers are circulated — while Pillar 1 scores near zero. Our rubric caps board reporting at the inventory score for exactly this reason: reporting quality cannot exceed the register it reports on.
Why it matters to the board
Oversight is discharged through information received and acted on. A board that received nothing has a documentation problem before it has a technology problem.
Board briefing → assessment → annual re-score
“Proprietary AI”
The pattern
A target presents AI as core intellectual property. Legal diligence checks licences and technical diligence checks the stack; neither asks what the model actually is, who owns the training data, or what the vendor terms do on a change of control.
What the Scorecard surfaces
Pillar 3 is where transactions turn: third-party dependency behind a proprietary label, training data whose provenance nobody can evidence, and key-person concentration where one engineer holds the undocumented knowledge.
Why it matters to the buyer
Each of these has a price. It is paid after close unless somebody looks before.
Five-day screen → deep review → 100-day integration plan
Every engagement begins with a confidential 30-minute briefing. Scope and fees are published.
Led by Ritesh Vajariya
Architect of BloombergGPT | CEO, AI Guru
Ritesh has built and deployed some of the most sophisticated enterprise AI systems in production. He drove $700M+ in annual AI/ML revenue at AWS, advised Fortune 500 leadership teams on AI deployment and risk, and architected core systems for BloombergGPT.
His work spans large-scale AI systems, enterprise governance, and real-world AI oversight.
NEUBoard brings that operational depth to the boardroom — helping directors govern AI with the same rigor they apply to financial controls, cybersecurity, and regulatory compliance.
Start the Conversation
Board-level AI oversight starts with the right questions. Let us help you ask them.
AI Guru is Ritesh's broader executive and enterprise AI platform focused on building and deploying AI systems. NEUBoard is a dedicated initiative focused exclusively on board-level oversight and governance.