Google's AI named Putin and Xi as dangers to the world but went soft on the US president invading Iran - and a German court has just ruled that whatever its AI says, Google said it.
The UN demands AI accountability while running Google ad-tech, analytics and tracking on its own website - with no consent and a measurable, avoidable carbon cost.
Amazon's Ring Familiar Faces biometrically scans every person who approaches a Ring doorbell, performs the match in Amazon's cloud rather than on the device that could have done it locally, and stores faceprints of non-consenting strangers for up to thirty days. The architectural choice produces two independent breaches: a GDPR exposure the household-activity defence cannot reach, and an avoidable datacentre carbon cost that brings the feature within the conceptual territory of the Environmental Crimes Directive. This piece walks through Articles 5, 6, 9, 13 and 25 GDPR, the Lindqvist, Rynes, Fashion ID and Jehovan todistajat line of CJEU authority, and Directive (EU) 2024/1203, and asks the regulators directly when biometric mass-processing of non-consenting people will be treated as the breach it plainly is.
Google has quietly removed the privacy assurance from Chrome's on-device AI Settings UI. The sentence promising that the model runs locally without sending data to Google's servers has been deleted, and the toggle moved out of the System block to reduce the chance the change is noticed. There are three plausible reasons for that, and each is a serious problem for users. This piece walks through the legal exposure under the EU Unfair Commercial Practices Directive, Section 5 of the FTC Act, and Articles 13(4) and 5(2) of the Digital Markets Act, and asks Parisa Tabriz directly why the assurance was withdrawn.
Google's Chrome boss Parisa Tabriz tells the press that users can simply opt out of the unsolicited Gemini Nano install. Google's own Chrome manifest proves the opposite. Chrome reached into the device, flipped the flag, downloaded the 4 GB model and only then surfaced the settings UI after the fact. Opt-out is not the legal standard here — opt-in is. This piece walks through why the public PR statements are demonstrably false against Google's own logs, and why a half-truth response to evidence is its own kind of harm.
Google Chrome is downloading a 4 GB Gemini Nano model onto users' machines without consent, with no opt-in, no opt-out short of enterprise tooling, and an automatic re-download every time the user deletes it. The pattern is identical to the Anthropic Claude Desktop case I wrote about last month, but the scale is between two and three orders of magnitude larger. This article does the legal analysis and, for the first time, the environmental analysis. The numbers are not small.
Bolt's customer support chatbot acknowledged a missing food item and twice refused a refund, swapping personas to dress automated denial as human review. A GDPR Article 22 case waiting to happen.
Anthropic ignored the Claude Desktop spyware findings. A formal Cease and Desist has now been issued, with 72 hours before criminal and civil complaints follow.
Anthropic's Claude Desktop silently installs a Native Messaging bridge into seven Chromium browsers, including browsers Anthropic's own documentation says it does not support, and browsers the user has not even installed.
