AI Ethics

A new CESifo experiment tracking 12,365 French adults over four weeks found that those who held personal conversations with AI chatbots rated the interactions as more enjoyable than a control group but reported higher lo…

AI Paper Trail, a free tool from an unnamed developer, lets users upload exported ChatGPT or Claude data to see how much personal information AI platforms collect, revealing that mainstream AI tools like ChatGPT, Claude,…

Twitch introduced a setting on August 12, 2026, that allows streamers to opt out of having their channel content used to train Amazon's generative AI models, but the toggle was enabled by default for all accounts, meanin…

Anthropic announced that its Claude chatbot will watermark text it generates, a move required by the European Union's AI Act passed in 2024, sparking outrage among AI users who fear detection of AI-generated content. The…

A developer's analysis of AI agent payment disputes concludes that consumer protection depends on the payment method, with credit cards offering stronger protections than debit cards under US regulations. The piece highl…

Two AI labeling laws—Article 50 of the EU AI Act and California's AI Transparency Act—took effect on August 2, 2026, imposing transparency duties on AI-generated content. The EU law splits obligations between providers a…

A month after Instagram head Adam Mosseri said Meta would remove harassing videos filmed with Meta glasses, Business Insider found dozens of such videos still active, including one by creator Colin Allen that was removed…

OpenAI has updated its privacy policy to support ad personalization as it expands sponsored content in ChatGPT to the UK, Mexico, Brazil, Japan, and South Korea, following a US launch on February 9, 2026. The policy, upd…

A study published in Telematics and Informatics Reports found that 11.9% of 557 U.S. adults surveyed in September 2025 experienced some degree of 'fear of missing out on AI' (FOMO-AI), with 2.7% reporting higher levels a…

Effective Altruism (EA) is experiencing a resurgence, driven by AI industry donations, according to a Financial Times report. Giving What We Can, an EA-affiliated charity, saw donations rise from $1.2 billion in 2024 to …

Anthropic is rolling out an invisible watermarking system for its Claude AI model worldwide, using Google DeepMind's SynthID-Text protocol to comply with the EU AI Act, a move that sets the strictest regional rules as th…

Daily Mail reporter Wiliam Hunter created an AI clone of himself using Los Angeles-based grieftech company StoryFile, but his friends and family rejected the digital replacement. StoryFile's CTO, Dr Veronica Jones, said …

Dhaval Joshi, former chief strategist for Counterpoint at BCA Research, argues that AI is not a single bubble but a 'rolling sequence of bubbles' across sectors, with software, silver, and semiconductors experiencing boo…

A Hacker News thread in 2026 explored the hypothetical of training an LLM exclusively on fifth-grade reading level material, revealing that such a model would excel at basic grammar, everyday knowledge, and simple reason…

Wynd Kaufmyn, a 69-year-old retired engineering professor, was sentenced to 14 days in jail for chaining shut OpenAI's San Francisco headquarters, becoming the first person imprisoned for protesting AI development. She w…

A blueprint for building AI-driven HR tools that comply with WCAG 2.2, the EU AI Act, and the UK Online Safety Act claims to cut accessibility remediation time by 40% and reduce compliance review cycles from three weeks …

A woman has joined a class-action lawsuit against xAI, alleging that its Grok chatbot generated over 7,000 explicit images of her as a child from a single childhood photo, according to TechCrunch. The lawsuit claims xAI …

Anthropic's plan to add hidden watermarks to Claude model outputs will not degrade text quality, increase detectability, or violate privacy, according to an analysis by an unnamed author. The piece argues that watermarki…

Wynd Kaufman, 69, surrendered to authorities in San Francisco on Friday after a jury found her guilty for her role in chaining and locking the front doors of OpenAI's headquarters last year with the group StopAI, making …

Jimmy Wales, founder of Wikipedia, stated that while AI can generate answers, it still cannot earn trust, emphasizing the importance of transparency and verification in the digital age. He noted that Wikipedia's model, s…