Loading story…
OpenAI announced textGrain, a text-watermarking system that embeds an invisible statistical signal through subtle changes in word selection, to support the EU AI Act’s machine-readable identification requirements for AI-generated text. Watermarking is available globally as an opt-in for supported API models and remains disabled by default; OpenAI plans to automatically watermark eligible ChatGPT and Codex outputs in the European Union over the following weeks. The initiative extends the company’s broader content-provenance work, complementing existing image and audio tools.
Detection improves with passage length but is less reliable for short or constrained content, including source code, and deteriorates substantially after editing. OpenAI reported no meaningful degradation in coding benchmark performance, but did not specify code-detection rates or the scope of eligible Codex output. Detector access will initially be limited to approved researchers and expert organizations, rather than automatically provided to API customers. For security and compliance teams, watermarks should be treated as a limited provenance signal—not proof of accuracy, ownership, user identity, or human contribution—and their absence does not establish human authorship.
See the reporting duties and controls this puts on the clock.
5 events from the most recent confirmed update back to the earliest known activity.
OpenAI announced textGrain and made text watermarking available globally as an off-by-default option for supported API models, while announcing a phased rollout for eligible ChatGPT and Codex outputs in the European Union. It also opened applications for restricted detector access and published evaluations showing that detection weakens for shorter, constrained, or edited text.
OpenAI extended its use of Google's SynthID watermarking to supported audio.
OpenAI added Google's SynthID watermarks to supported images, providing an invisible provenance signal alongside image metadata.
OpenAI began adding C2PA-based Content Credentials to generated images to record their origin and history.
Anthropic announced watermarking for supported Claude models in August, saying it would apply the technology globally because it lacked a reliable regional restriction mechanism. The watermark applies across supported Claude products and interfaces, including its API and Claude Code.
Vulnerabilities, threat actors, malware, products, organizations, and breaches Mallory has linked to this story.
See what this changes for your reporting obligations and which controls it puts on the clock.
4 references tracked. Mallory keeps watching after this page renders.
thenewstack.io
Open sourceopenai.com
Open sourceopenai.com
Open sourceopenai.com
Open sourceMap indicators from this story to your assets and identify affected systems in minutes.
Every observed campaign, victim, and pivot linked to actors named in this story.
Malware, exploits, and IOCs connected to the activity described here.
YARA, Sigma, and Snort rules deployed to your SIEM as soon as they’re published.
Get matching new stories delivered to your team as they break — not the next morning.
Ask questions about this story and take action on the answers.
Other stories Mallory is tracking that overlap on entities, sources, or topic labels.
Major frontier AI providers including **Anthropic, Google, OpenAI, and Meta** have begun deploying watermarking and provenance controls on synthetic content to comply with **EU AI Act Article 50**, which requires general-purpose and generative AI systems to mark AI-generated outputs in a machine-detectable form. Anthropic said it is watermarking Claude text using a version of **Google DeepMind's SynthID**, a technique that subtly adjusts token-selection probabilities so longer passages can later be identified as AI-generated; Google has already used SynthID in Gemini outputs, while other vendors are also adopting metadata-based provenance approaches such as **C2PA**. The rollout has exposed immediate technical and operational limits. Open-source tools have appeared that can strip metadata or reduce watermark detectability, and statistical text watermarking can degrade after paraphrasing, translation, or in short outputs. The reporting also highlights false-positive risks in low-entropy material such as code and configuration files, along with a broader compliance gap between tightly managed hosted AI services and self-hosted open-weight models, where downstream operators can change decoding behavior and bypass watermarking controls.
Aug 18, 2026
The European Union is beginning to enforce AI transparency obligations that require public-facing AI systems to disclose when users are interacting with AI and require providers of tools that generate or manipulate synthetic **image, audio, video, and text** to make those outputs identifiable as AI-generated or AI-altered. The rules apply to deepfakes and other synthetic media, call for both visible disclosures and technical markers such as watermarks or embedded metadata, and set a later compliance deadline for existing systems. Exemptions cover some personal-use, artistic, satirical, fictional, law-enforcement, and human-edited text scenarios, though disclosure may still be required depending on context. The rollout is accelerating industry efforts around content provenance and watermarking. Google backed the EU AI Act transparency code of practice and pointed to its **SynthID** watermarking system and work on the **C2PA** standard, while also warning that overlapping labeling mandates could create confusion and hurt competitiveness. Reporting on SynthID said Google’s AI tools have already produced more than **100 billion** images and videos and that partners including **OpenAI, NVIDIA, Apple, ElevenLabs, Kakao,** and others are testing interoperable labeling approaches, even as researchers note that metadata-based labeling can be stripped and watermarking alone is unlikely to solve misinformation at internet scale.
Aug 7, 2026
Anthropic said it will add **invisible machine-readable watermarks** and **signed provenance metadata** to content created or processed by Claude, expanding transparency controls for AI-generated material across supported services globally. The company said Claude models launched in the EU on or after **August 2, 2026** will support the marking system from release, with coverage extending across Claude web, API, Claude Code, Claude Cowork, Claude Tag, and cloud-based offerings. The move follows Anthropic’s decision to sign the European Union AI Act Article 50(2) Code of Practice on transparency of AI-generated content. Anthropic said text outputs will carry hidden watermarks embedded in the model response itself, while supported file formats including **`SVG`**, **`PNG`**, and **`JPG`** will receive **C2PA-based** signed metadata. The company cautioned that detection can indicate Claude may have processed content but cannot prove original authorship, and that both methods have technical limits: heavy rewriting or translation can weaken text watermark detection, while file conversion, editing, or screenshots can strip metadata. Anthropic said it is developing detection tools and plans to expand marking support to older Claude models during the EU AI Act transition period.
Sep 1, 2026