Thought leadership on ethical AI, peacebuilding technology, and democratic innovation
Ai Governance Oversight
Treating AI oversight as a box-ticking exercise wastes its real value: sustained confidence in systems whose behaviour keeps changing long after sign-off. Compliance proves a system was acceptable once, while confidence tells you it still is today, which is a different and considerably harder question.
Ai Governance Oversight
Deployment is where AI governance starts working, not where it ends, because models drift, data shifts and controls decay once a system meets the real world. Sign-off is a checkpoint rather than a finish line, and treating it as one is how quiet failures accumulate until they are expensive.
Ai Governance Oversight
Governance is usually experienced as a brake on delivery, which is precisely why teams route around it. Treating oversight as a product feature, with named owners, clear interfaces and tests, makes it something that ships alongside the model rather than something imposed on it afterwards.
Dive deeper into our key research and practice areas.
How false information spreads and why it works on us - and the critical thinking, prebunking and resilience strategies that help individuals, organisations and democracies push back.
19 articles
How platforms shape public discourse through algorithms and moderation, and how liability rules and regulation like the Digital Services Act are trying to hold them to account.
4 articles
AI regulation and digital policy across the UK, EU and beyond - what the emerging frameworks demand of organisations, and the rights they exist to protect.
4 articles
Technology in humanitarian contexts: protecting vulnerable people, supporting crisis response, and confronting the harms AI can inflict on refugees and marginalised communities.
1 article
Defending democratic processes from influence operations, voter suppression narratives and synthetic media in an AI-shaped information environment.
4 articles
Practical AI governance: oversight frameworks, audit trails, drift monitoring and risk management that keep deployed AI systems accountable from design through to live operation.
23 articles
Deepfakes and generative AI content - what synthetic media means for consent, trust, law and culture, and the struggle to preserve authenticity online.
33 articles
Intelligent systems in war and peacebuilding: early warning, machine-mediated reconciliation, autonomous weapons, and AI's role in societies rebuilding after conflict.
27 articles