Description

AI is not just another tool rollout. It is forcing companies to revisit assumptions they have lived with for years.

In this Risk Grustlers episode, Sandip Wadje joins Nicholas Muy to discuss what happens when leadership wants the upside of AI immediately, while security, risk, IT, legal, and business teams are left dealing with everything underneath it.

The conversation explores what tools like Microsoft Copilot can expose inside organizations, from oversharing in SharePoint to the gap between role-based access on paper and the permissions people actually end up with in practice.

Sandip also lays out a more grounded approach for smaller teams: be honest about what is truly confidential, avoid writing controls you cannot consistently follow, give employees a safer way to work with AI, and start preparing for scrutiny around how AI is being used across the business.

Key AI governance insights from this episode

• Why does data classification suddenly become urgent when AI begins inheriting access across existing systems
• How AI exposes entitlement problems that traditional role reviews often fail to catch
• What smaller organizations should do first, instead of copying heavyweight AI control programs built for large enterprises

Quote from the Episode

“Don’t write internal controls that you would not be able to follow.”
‍
— Sandip Wadje, Managing Director, BNP Paribas