Sep 26, 2026
Texas District Judge Cory Liu ruled Thursday that TikTok violated the state's consumer protection laws by misleading the public about its ability to filter inappropriate and sexually-explicit content for minor users. The ruling found that TikTok's claims about its content-moderation policies and 'Restricted Mode' feature were deceptive.
Instagram announced it will begin restricting the reach of accounts featuring AI-generated people that fail to disclose their artificial nature. The platform is introducing an 'AI-generated profile' label and renaming its existing AI creator label as part of new transparency measures.
OpenAI shut down a cluster of ChatGPT accounts originating in Russia that were used to create content for a fake Israeli think tank called the International Burke Institute. The operation used VPNs and Russian-language prompts to generate English social media posts promoting pro-Kremlin narratives and fabricated academic research.
A new Pew Research study analyzing roughly 490,000 English-language webpages found that over one-third of pages published after ChatGPT's November 2022 release show significant signs of being written by AI. The findings, based on Common Crawl archive data through July 2026, corroborate other recent studies on AI-generated web content.
Billionaire investor Stanley Druckenmiller disclosed that he used artificial intelligence to write a Wall Street Journal op-ed attacking Treasury Secretary Scott Bessent's debt buyback policies. The admission comes amid growing concerns about AI-generated content and the difficulty of detecting it.
Research reveals that AI systems tend to favor AI-generated text over human-written content, while experts warn that tools designed to detect AI-generated text are fundamentally flawed and unreliable.
Within hours of Anthropic confirming the global rollout of invisible watermarks in Claude-generated text, developers including Guillaume Meyer published tools and workarounds to remove or override the watermarking system. The rapid response demonstrates immediate pushback against AI content labeling efforts.
Anthropic announced it will watermark text from its Claude chatbot, while the EU begins enforcing transparency requirements for AI-generated content. The moves come as publishers like the New York Post deploy AI tools and concerns grow about deepfakes.
Google announced it will let users toggle off visible watermarks on images, videos, and music created by its Gemini AI assistant. The company will continue embedding invisible SynthID and C2PA watermarks for content authentication.
Anthropic is implementing invisible watermarks on all text and file outputs from its Claude chatbot worldwide to meet EU AI Act transparency requirements, effective August 2. The move has drawn complaints from some users concerned about detection of AI use in work and academic settings.
Anthropic announced a watermark detection API allowing third parties to verify Claude-generated text, implementing the feature globally to comply with EU AI Act requirements. However, researchers discovered the watermark can be stripped for approximately four cents per attempt, as the public API inadvertently functions as an evasion oracle.
Anthropic announced it will add machine-readable watermarks to text and files generated by future Claude models starting August 2, 2026, in response to transparency requirements under Article 50 of the EU AI Act. The watermarking will apply globally to all users with no opt-out option.
Suno announced it will implement invisible audio watermarks on AI-generated tracks and impose new download limits to prevent mass distribution, as the company faces copyright infringement lawsuits from major music labels including Sony and Universal.
French AI startup Mistral AI announced Shieldstral on August 4, 2026, a 3-billion parameter open-weights multimodal safety classifier that matches models up to seven times its size in performance while allowing operators to define custom moderation policies in natural language at runtime.
Article 50 of the EU AI Act entered into force, requiring AI providers and deployers to label chatbots and AI-generated deepfakes. The new transparency obligations apply to all generative AI tools operating in the European Union.
Snapchat announced it will no longer recommend or reward entirely AI-generated videos on its Spotlight feature, prioritizing human-made and AI-assisted content created with its own tools instead. The policy change took effect in July 2026.
LinkedIn introduced new ways to reduce AI-generated posts, including a 'seems like AI slop' reporting option, and replaced its AI writing feature with a proofreading tool. Snapchat also announced it will no longer recommend or reward fully AI-generated videos in Spotlight.
Google rolled back its new AI feature that allowed users to generate artificial intelligence images superimposed on satellite maps after BBC Verify reported it could create realistic deepfakes of bombings, riots, and destruction. The feature used Google's Nano Banana 2 model.
Australia's eSafety Commissioner filed civil penalty proceedings against Telegram seeking up to A$54.6 million ($38 million), alleging the platform failed to comply with removal notices for terrorism-related content.
A study from a think tank led by former FTC Chair Lina Khan reveals that Amazon and Walmart's AI shopping assistants can identify fraudulent 'Made in USA' product labels but the retailers are not deploying this technology to remove or flag such listings.
A study from a think tank led by former FTC Chair Lina Khan reveals that Amazon and Walmart's AI shopping assistants can identify fraudulent 'Made in USA' product labels but the retailers are not deploying this technology to remove or flag such listings.
Meta introduced Content Seal, an invisible watermarking system for AI-generated images from its Muse models, following pressure from its Oversight Board to combat deceptive AI content. Early tests reveal significant detection gaps, with critics questioning why Meta didn't adopt Google's existing SynthID technology instead.
Patreon has shifted from using robots.txt files to request AI bots not scrape creator content to actively blocking unauthorized AI training crawlers through Cloudflare's Crawl Control technology. The platform reports the new enforcement approach has significantly reduced scraping activity while still allowing legitimate search and discovery bots.
San Francisco City Attorney David Chiu sent cease-and-desist letters demanding Apple and Google remove 13 apps that generate non-consensual nude images. The apps can be used to create deepfake nude images of people without their consent.
Cloudflare announced it will block AI training crawlers by default starting September 15, 2026, requiring AI companies to separate their search crawlers from training bots. Pages displaying ads will automatically block training and agent bots unless publishers explicitly allow access, forcing AI companies to negotiate payment for content.
Canva's AI feature was discovered automatically replacing the word 'Palestine' in user designs, prompting an apology from the company. The incident raises concerns about content moderation in AI-powered design tools.