Apple's iOS 27 beta 5 includes a hidden 'Apple Reference Image' feature that will authenticate photos and videos captured by an iPhone, using metadata and Private Cloud Compute to verify they are uned…
Anthropic's distributional text watermark is carried by word choice, not invisible characters, so tools that strip zero-width Unicode characters cannot remove it, according to Claude Watermark Researc…
Anthropic rolled out an imperceptible watermark for AI-generated text in its Claude models, effective August 2, as part of its commitments under the European Union AI Act. The watermark, which persist…
Article 50 of the EU AI Act became enforceable on August 2, 2026, imposing transparency obligations on AI systems that interact with users or generate synthetic content. The regulation requires machin…
Anthropic will embed machine-readable watermarks into all text generated by Claude models launched after August 2, 2026, complying with the EU AI Act's Article 50(2) Code of Practice. The watermarks, …
Apple is developing a new feature called Apple Reference Image, found in iOS 27 beta 5, to authenticate that a photo was captured by an iPhone camera using unique hardware data. The system sends the i…
OpenAI launched GPT-Live on July 8, introducing GPT-Live-1 for paid subscribers and GPT-Live-1 mini for free users globally, while Google expanded its Gemini 3.5 Live Translate feature since June, bot…
Suno, the AI music generation company, announced it will add watermarks to all audio outputs from its models to meet emerging industry standards for labeling AI content, according to a blog post by CE…
OpenAI has expanded its content provenance work with SynthID watermarking for audio from GPT-Live in ChatGPT Voice and the OpenAI API, plus a verification API for checking whether a supported audio fi…
A daily AI news digest published on Aug 3 lists 23 stories, including Amazon's $50 billion investment in OpenAI, Reddit's CEO questioning Google AI Overviews amid stock declines, and an AI COBOL-to-Ja…
DeepMind's SynthID watermarking technique, while cryptographically robust, fails to solve AI disinformation because it is not a universal standard, can be evaded by paraphrasing, and only indicates co…
Google is rolling back a new feature in Google Earth that allowed users to generate AI visualizations on satellite imagery, following backlash from researchers and open-source intelligence experts who…
A security researcher has published a detailed attempt to extract Google DeepMind's SynthID image watermark detector and explore adversarial machine learning attacks against it, revealing that the inv…
Google pulled its Google Earth AI image generator within 24 hours of its July 30, 2026 launch after researchers demonstrated it could fabricate photorealistic images of bomb craters near the Kremlin, …
Google launched a new feature in Google Earth on Thursday that lets users generate AI-powered visualizations layered over satellite imagery using its Nano Banana 2 image-generation technology, sparkin…
The Verge reports that Google has rolled back its AI image generator in Google Earth after it was used to create realistic satellite images of a bomb crater near a hospital in Gaza, which could be mis…
Google removed an AI image generation feature from Google Earth after users, including open source intelligence researcher Hank Van Ess, used it to create fake satellite images depicting war zones and…
Google's Gemini-powered image generator Nano Banana in Google Earth can create photorealistic composites of real map data with no visible marker distinguishing synthetic overlays from real imagery, a …
Google launched a Google Earth web feature on July 30 that lets users generate images from satellite, aerial, and 3D map views using a text prompt, but NPR, AFP, and open-source researchers documented…
Google's new Nano Banana AI image generator in Google Earth can create realistic satellite and aerial images that are ripe for misuse, as demonstrated by Digital Digging's Henk van Ess, who generated …