Reuters: software companies are repositioning products and business models to stay relevant as generative AI changes buyer expectations and competition dynamics.

Reuters: the Pentagon’s CTO said there is “no chance” of renewed negotiations with Anthropic, underscoring how procurement, reputational risk, and model-use constraints are now governance levers.

People: YouTube is expanding a programme that lets journalists and political officials request removal of AI deepfakes impersonating them, an operational shift in platform governance against synthetic identity abuse.

Regulation

  • UK Government (BEIS/DSIT via GOV.UK): updated national security investment rules refinement states “off‑the‑shelf” AI systems will be removed from mandatory notification, with focus shifting to firms developing or modifying advanced AI.

  • UK Government (DSIT) privacy notice: follow‑up survey activity relating to the EU Cyber Resilience Act and EU AI Act (page published 12 March 2026).

  • Ofqual (GOV.UK guidance): “AI in coursework: resources for schools” (published 9 March 2026), practical governance for assessed work and AI use in education settings.

Academia

  • arXiv: Engineering a Governance‑Aware AI Sandbox: Design, Implementation, and Lessons Learned (posted 3 March 2026), operationalises governance constraints as system features rather than policy overlays.

  • arXiv: Governance of AI‑Generated Content: A Case Study on Social Media (posted March 2026), examines platform governance mechanisms for AI content at scale.

  • SSRN: What Rights Do AI Companies Have in Government Procurement? (Date written 10 March 2026), timely for vendor policy restrictions, exclusions, and procurement accountability.

Events

  • UN Indico: STEPAN Webinar Series 2026 – Advancing Responsible AI: Readiness and Governance (virtual, 12 March 2026).

  • OneTrust: Leading the Charge on Trustworthy AI Governance (webinar, 25 March 2026).

  • EU IP SME Helpdesk / FinnCham China: Beyond the EU AI Act – Protecting Trade Secrets and Structuring IP in China’s AI Era (webinar, 9 April 2026).

Takeaway

Governance is moving from abstract principles to concrete control points: (1) state screening rules that define what counts as advanced AI for investment scrutiny, (2) platform mechanisms for identity/deepfake enforcement, and (3) procurement disputes where vendor-use restrictions collide with government deployment aims.

Sources: GOV.UK; UN Indico; Reuters; People; OneTrust; EU IP SME Helpdesk; arXiv; SSRN