GoI has directed Meta to disable ads and content on Instagram that reportedly facilitate child sexual abuse material. This was after a BBC report found the social networking service running paid ads of that nature in India. Meta responded by stating that it uses 'advanced AI technology to proactively detect violating content and individuals', adding that it was 'categorically inaccurate' to suggest it knowingly targeted such ads to users, and denied prioritising revenue over safety.
Meanwhile, in an FT article earlier this month, Sam Altman called for governments - not companies - to set rules for AI. The next day, OpenAI was reported to have discussions on giving the US government 5% stake in the company. Whenever incumbents begin speaking passionately about 'global standards', 'responsible governance' and 'shared safety', history suggests they have usually reached the stage where they own most of the assets and would like everyone else to stop catching up. It was nuclear clubs earlier, followed by semiconductor alliances. Now, it's AI safety.
