AI rules are splintering
Europe is preparing a ban on non‑consensual sexual deepfakes after recent Grok‑linked scandals, while regulators worldwide are imposing wildly different requirements — from China’s two‑hour transparency updates to a patchwork of state rules in the US reported and summarized. Businesses say the confusion is real — surveys flag uneven implementation and liability risks — even as companies deploy AI to stabilize supply chains and consumer products like Amazon’s new “Sassy” Alexa mode roll out with added privacy controls argued and reported.
EU ambassadors backed ((france24.com)) a proposal on March 13, 2026 to prohibit AI systems that generate sexualised, non‑consensual deepfakes, and the move was explicitly tied in those debates to images made with Elon Musk’s Grok ((france24.com)). X.AI’s Grok has prompted multiple national probes and a US class action alleging failure to prevent thousands of non‑consensual images, with the complaint filed in January 2026 against X.AI citing prolific creation and spread of explicit deepfakes ((news.bloomberglaw.com)). China’s cyberspace regulator issued draft rules on Dec. 27, 2025 that would force providers to identify AI at login and again at two‑hour intervals for human‑like services, and the draft opened for public comment into January 2026 ((bloomberg.com)). State‑level US activity produced more than 1,100 AI bills in 2025 and a US Chamber analysis finds 65% of small businesses worried about compliance and litigation from a patchwork of state rules, while the federal executive order seeks to preempt those diverse laws ((uschamber.com)). Surveys show governance gaps: PwC’s responsible‑AI polling found about half of executives cite operationalizing controls as the main hurdle, and Moody’s 2025 risk survey polled roughly 600 compliance professionals who flagged monitoring and liability as top concerns ((pwc.com)). Corporate deployments are already splitting the picture—Amazon described three AI supply‑chain innovations including a generative mapping tool and an AI demand‑forecasting model in 2026 ((aboutamazon.com)), and Amazon’s Alexa+ launched a new adults‑only “Sassy” personality with opt‑in security checks (including Face ID on iOS) and limits to prevent sexual content or minors’ exposure ((aboutamazon.com)).
