@lordtim.bsky.social

Who in government owns AI safety? As of this week, nobody can say for certain. Tim Clement-Jones, Lib Dem science, innovation and technology spokesperson, forced ministers to answer that question in the Lords today.

We need to introduce Algorithmic Impact Assessments for recruitment and hiring algorithms, so that employers and platforms are required to test for bias

There are growing threats to democratic institutions in the United Kingdom as we see the collapse of public trust. We need cross-sector AI regulation, including mandatory watermarking of synthetic content and to seriously invest in digital and media literacy.

Bild

Given the UK security vulnerabilities demonstrated by models like Mythos the government should bring forward legislation giving the AI Security Institute the statutory power to require, and where necessary delay launch of dangerous frontier AI models

Given the UK security vulnerabilities demonstrated by models like Mythos the government should bring forward legislation giving the AI Security Institute the statutory power to require, and where necessary delay launch of dangerous frontier AI models

The Prime Minister seems finally to have U turned on social media harms for the under 16's. Now we need action, with a film-style harms-based age-rating system, with addictiveness as a central criterion, rather than a blanket ban on user-to-user services.