AI Oversight Meets Evidence

Reuters reported that South Africa withdrew its first draft national AI policy after fictitious sources appeared in its reference list. The draft had proposed a National AI Commission, an AI Ethics Board and an AI Regulatory Authority, but the withdrawal put human review and citation integrity at the centre of policy credibility. 

Reuters reported that Microsoft will no longer have exclusive access to OpenAI models and products. The revised arrangement leaves Microsoft as OpenAI’s primary cloud partner while allowing OpenAI to sell across rival cloud platforms, including Amazon and Google. 

Regulation

Ofqual published an advice note on artificial intelligence malpractice and assessment. The document explains how existing Conditions of Recognition and guidance apply to learner use of AI tools, including reasonable steps for awarding organisations to identify and manage assessment malpractice risks. 

Academia

SSRN hosted ‘EU AI Act: Obligation-to-Control (O2C) Framework’. The paper is relevant to AI governance because it treats the EU AI Act as a control architecture and links compliance duties to operational oversight rather than paper-based assurance alone. 

Events

IAPP will hold the Canada Symposium 2026 in Toronto from 4 to 7 May. The event covers privacy, AI governance and cybersecurity law, with a Canadian and global focus on digital responsibility practice.

Digital Hollywood will hold AI Legislation and the Law 2026 in Washington DC on 5 and 6 May. The event focuses on AI legislation, technology and law, including state AI regulation, pending federal activity and AI lawsuits. 

Sources: Reuters, Ofqual, CourtListener, SSRN, IAPP, Digital Hollywood