UK privacy watchdog secures AI company commitments on how personal data is used to train AI models

What happened

On Oct. 8, 2026, the UK Information Commissioner's Office (ICO) announced that 10 major AI developers have made or committed to make changes following its two-year supervisory programme examining UK data-protection compliance across AI training practices. The named companies — Google, Microsoft, Amazon, Apple, Anthropic, OpenAI, Meta, Cohere, DeepSeek and Stability AI — committed to clearer explanations of how personal information is used to train AI models, stronger data-rights mechanisms, and tougher safeguard assessments. The ICO said it is still monitoring delivery of the commitments, opened a six-week call for evidence on agentic AI (due Nov. 20), and moved xAI's Grok into a separate formal investigation.

Why it may matter to you

This is the first multi-company, binding-style regulatory outcome on training-data practices to land in the corpus: it sets a UK compliance baseline for how personal data feeds model training and gives creators and rights holders a concrete reference point (clearer disclosures, stronger data rights) for challenging unconsented training use. The separate formal investigation into xAI's Grok is the sharp end to watch.

What is not yet known

Which commitments each company has actually delivered (the ICO is still monitoring), what enforcement action — if any — the separate Grok investigation produces, and whether other regulators (EU, US) follow with parallel probes.