DeepMind Announces AGI Breakthrough, Immediately Classifies All Details

LONDON — DeepMind issued a press release Tuesday afternoon confirming that it had achieved artificial general intelligence, describing the development as 'a significant scientific milestone' and 'the culmination of decades of foundational research.' The release was 47 words long. All subsequent requests for elaboration have been referred to the company's legal department, which has not responded.

The announcement arrived without warning, without a research paper, and without the customary pre-briefing of science journalists. DeepMind's communications team confirmed the release was intentional and said a 'fuller disclosure framework' was being developed. Asked when that framework would be ready, a spokesperson said she was not able to share that information at this time.

The AI research community responded with a mixture of excitement and frustration. Several prominent researchers noted that 'AGI' has no universally agreed definition, making the claim difficult to evaluate without supporting data. Others pointed out that DeepMind has historically published detailed technical work and that the sudden opacity was itself a significant signal. 'They've never gone quiet before,' said one researcher at a competing lab, who asked not to be named. 'That tells you something.'

Alphabet, DeepMind's parent company, saw its share price rise 34% in after-hours trading before regulators halted the stock pending clarification of the announcement. The clarification has not been provided. Alphabet's CEO issued a statement saying the company was 'committed to responsible disclosure' and would share more information 'when the time is right.' The statement was 31 words long.

As of publication, DeepMind's website has been updated to remove all references to its previous AI safety research roadmap. The company's research blog has not been updated since Monday. A cached version of the blog's final post, published at 11:47 PM Sunday, ends mid-sentence.