As another AI researcher quits, warning of an existential risk to humanity, industry insiders demand enforceable treaties and a global licensing regime for superintelligence – before it’s too late.

Whistleblower Tumiso Mphuthi exposed corruption, was vindicated by subsequent investigations, then fired anyway. South Africa’s protections remain “performative political theatre”.

Google is moving its 90-person AI responsibility team out of DeepMind into global affairs, fuelling fears that the shift will compromise access, independence and threat-detection ability.

Professor Egbuna Olakunle Obidike, the architect of Gambian paediatrics whose training programme caught the toxic cough syrup poisoning children worldwide, has passed away. Tributes are pouring in.

Europe’s AI Act enters enforcement with nine banned practices, high-risk safeguards, mandatory watermarking, clearly labelled deepfakes and fines for non-compliance of up to 3% of global turnover.

South Australia will hold Australia's first royal commission into AI – not to investigate a past failure, but to shape policy before the damage is done.

State control of the news media is shaping what AI chatbots say. Research in Nature finds LLMs give more pro-government answers in the languages of countries with less press freedom.

‍ ‍

Anthropic secretly scanned and pulped millions of books to train its AI, because pre-2022 print is free of AI-generated slop now degrading large language models.

Bernie Sanders wants citizens to own half of Big AI. Critics warn public stakes would make governments protect the firms they should be regulating, amounting to capture, not accountability.

All Ricochet Refresh users must update to versions newer than 3.0.43 (released June 2, 2026) before September 1, 2026, when a Tor protocol change breaks older clients’ network access.

Nigeria's courts have prioritised pre-election disputes over a whistleblower's freedom. Nnamdi Emeh's trial has been marred by endless delays in a case that raises profound questions about the respect for due process and the rule of law.

A landmark ruling could herald a new era of legal liability for fake information generated by AI tools