Creating a glass box: How NetSuite is engineering trust into AI

Source: VentureBeat

Overview

Presented by Oracle NetSuite

When any company tells you it is their biggest product release in almost three decades, it’s worth listening. When the person saying it founded the world’s first cloud computing company, it’s time to take note.

At SuiteWorld 2025, Evan Goldberg, founder and EVP of Oracle NetSuite, announced the launch of NetSuite GlassBox, a suite of AI‑driven tools designed to embed transparency, governance, and compliance directly into NetSuite’s AI capabilities.

Why “GlassBox”?

• Transparency: Provides real‑time visibility into how AI models make decisions, including data lineage and feature importance.
• Governance: Enforces policy controls, audit trails, and role‑based access to AI outputs.
• Compliance: Helps organizations meet regulatory requirements such as GDPR, CCPA, and industry‑specific standards.

Core Components

1. Explainable AI (XAI) Dashboard – Visualizes model reasoning, highlights influential data points, and offers “why‑this‑result” explanations for end users.
2. Policy Engine – Allows administrators to define rules (e.g., “do not use personal data for credit scoring”) that are automatically enforced during model inference.
3. Audit Log – Captures every AI interaction, including input data, model version, and output, creating a tamper‑proof record for auditors.
4. Model Registry – Central repository for all trained models, with version control, metadata, and lifecycle management.

Integration with NetSuite ERP

GlassBox is built natively into NetSuite’s ERP suite, meaning:

• AI‑enhanced financial forecasting, inventory planning, and demand sensing can now be inspected and governed without leaving the platform.
• Existing NetSuite customers can adopt GlassBox features through a simple configuration update—no separate infrastructure or third‑party tools required.

Security & Data Privacy

• Data Residency Controls – Customers can restrict model training and inference to specific geographic regions.
• Differential Privacy – Optional noise injection to protect individual records while preserving aggregate insights.
• Encryption‑in‑Transit & At‑Rest – All AI payloads are secured using industry‑standard TLS and AES‑256 encryption.

Industry Impact

Goldberg emphasized that “trust is the new currency for AI.” By providing a “glass box” rather than a “black box,” NetSuite aims to:

• Reduce risk of biased or non‑compliant AI decisions.
• Accelerate AI adoption across finance, supply chain, and HR functions.
• Differentiate NetSuite in a crowded ERP market where AI capabilities are becoming a baseline expectation.

Availability

• Beta Launch: Early 2025 for select enterprise customers.
• General Availability: Q3 2025, with tiered pricing based on usage and feature set.

What’s Next?

NetSuite plans to expand GlassBox with:

• Auto‑ML Pipelines – Guided model creation with built‑in compliance checks.
• Third‑Party Model Marketplace – Curated AI models vetted for transparency and governance.
• Enhanced Collaboration Tools – Shared workspaces for data scientists, auditors, and business users to co‑manage AI projects.

---

For more details, visit the official NetSuite announcement page or contact your NetSuite account representative.