Category: Governance

AI governance frameworks, oversight models, accountability structures, and organisational control over AI use

Meaningful Human Involvement Is Not A Rubber Stamp Europe’s First Uber Robotaxi Still Has A Safety Operator France excludes OpenAI from cyber work Ten Days To Answer A European Production Order UK growth: AI vendors, not AI users Two clouds. One jurisdiction. No regulator paused Astra. OpenAI did. US productivity up 1.4%, cause unproven Moldova’s GDPR law lands on 23 August

Stay ahead of AI change. Get practical updates from Future Prep direct to your inbox.

By subscribing you agree to receive Future Prep news. Unsubscribe any time.

Glass distillation apparatus in a workshop with a row of derivative receiving vessels behind it, illustrating model provenance and the difference between an original AI system and its distilled copies.

Model Provenance: 3 Critical Vendor Questions

A US State Department cable on Chinese AI distillation has turned model provenance into an immediate vendor due diligence question for EU deployers. Three concrete asks for every supplier this quarter, and where the EU AI Act helps but does not finish the job.
Industrial scale weighing EU-flagged servers against an unmarked unit illustrating the cloud sovereignty framework procurement assessment

SEAL-2, SEAL-3 and the Architecture of the EU’s €180 Million Sovereignty Bet

The European Commission awarded €180 million in sovereign cloud contracts assessed against its new Cloud Sovereignty Framework and SEAL levels. A working reference for procurement teams.
Postcard and sealed envelope on a park bench illustrating AI acceptable use policy confidentiality gap

When AI Chats Become Evidence: What the GWG Holdings Ruling Means for Your Acceptable Use Policy

A US court ruled that AI chatbot interactions are not privileged and can be compelled as evidence. Most organisations' AI acceptable use policies are not ready for this. Here is what to fix.
Tree splitting a boundary wall between two contrasting landscapes representing AI provider jurisdiction risk

Your AI Provider Just Got Blacklisted. Now What?

When your AI provider gets blacklisted in one country and courted by another, governance teams face a new risk category: AI provider jurisdiction risk.
Split desk scene representing AI data-sharing readiness between government and AI provider

When Governments Want Your AI Data

The Australia-Anthropic MoU signals a new regulatory pattern: governments negotiating direct access to AI usage data. EU governance teams should prepare now.
A hand hovering over a payment terminal beside a tablet showing abstract analytics, representing the governance question of who authorises action when AI systems can execute decisions.

From chatbot to agent: what changes when AI starts doing, not just saying

Starling Bank's AI moves money. Norway's sovereign wealth fund keeps humans in charge. Both are deploying agentic AI but with very different governance choices. Here is what changes when AI starts acting, not just recommending, and what your oversight framework must address.
Scroll to Top