AI Ethics news and analysis on Web Pulse: 16185 curated articles tracking the latest AI Ethics developments, tools, and research, updated continuously from vetted sources.
Twitch has introduced a "Training for Generative AI" control in channel settings, enabled by default, that allows streamers to opt out of having their content used to train Amazon's generative AI models. Chief product of…
Twitch is now using stream data to train generative AI models, according to a report. The platform frames this as a way to build better creator tools, but creators' unique voices and community interactions are being inco…
AI crawlers from Meta and Alibaba nearly took down the LGBT History Project, a volunteer-run archive founded by Jonathan Harborne, after one Meta crawler made roughly 26,000 requests in a day and pulled about 12GB of dat…
AI governance debt is compounding as enterprises prioritize speed-to-market, with agentic AI systems scaling errors at machine speed, according to a report citing IBM research that AI portfolio returns drop by 18% to 29%…
Twitch confirmed on Wednesday that it is using channel content to train Amazon's AI models, with the feature automatically enabled unless creators opt out via a new setting. The company added an opt-out button to account…
Meta is donating 15,000 Ray-Ban Meta smart glasses to the National Council for the Blind of Ireland (NCBI), a charity supporting people with sight loss in Ireland, to help blind and visually impaired individuals access A…
Enterprises integrating probabilistic AI systems must adopt AI Validation Engineering, a discipline that goes beyond traditional functional testing, according to a new framework outlining five validation patterns. The pa…
NVIDIA signed a memorandum of understanding with Apollo Global Management, BlackRock, Blackstone, Brookfield Asset Management, Goldman Sachs, and KKR to build a compute financing platform targeting over $500 billion in t…
Twitch, the Amazon-owned livestreaming platform, has enabled a default setting that allows channel content to train generative-AI models at Amazon, with an opt-out control that users began circulating on August 12, 2026.…
Apple Inc. is in talks with news publishers to license content for its AI-powered Siri assistant, according to the Wall Street Journal, despite a history of AI-generated news summaries producing false claims. In late 202…
Japanese companies are falling behind in AI adoption due to a consensus-based decision-making process (ringi) that conflicts with AI's need for rapid experimentation, according to an analysis of cultural and structural b…
Twitch, a subsidiary of Amazon, has enabled by default a new setting that allows user channel content to train Amazon's generative AI content models, with users able to opt out via the Security and Privacy tab. Twitch Ch…
Software engineers are rediscovering the importance of craftsmanship as AI tools like large language models threaten to automate coding, according to a personal essay by a research software engineer. The author argues th…
Gusto, a human resources platform, has introduced a five-tier AI archetype system for quarterly employee performance reviews, ranging from 'observers' to 'amplifiers,' to standardize AI fluency discussions, according to …
More than 120,000 tech jobs have been cut globally in 2026, according to Layoffs.fyi, and a survey by Lenny's Newsletter of 8,200 workers found burnout rose from 44.4% to 55.7% and career optimism fell from 54.8% to 48.7…
A developer detailed the steps taken to harden a vendor's agent runtime before deployment, specifically for Quark Drive's MCP server package. The developer disabled telemetry, self-updates, and other features by patching…
An engineer explains why large language models hallucinate, attributing the issue to statistical prediction rather than factual knowledge, and outlines practical mitigation strategies such as grounding models with source…
Gracenote, a provider of entertainment metadata, reports that ungrounded large language models (LLMs) hallucinate details for nearly 1 in 5 movie and TV titles tested, citing an example where an LLM confused Taylor Sheri…
Twitch's Chief Product Officer admitted during a public exchange that the platform's new AI training feature is widely disliked, acknowledging that if it were opt-in, nobody would choose it, yet the company plans to proc…
His Majesty’s Courts and Tribunals Service (HMCTS) has banned Meta smart glasses from courts and tribunals in England and Wales, with devices confiscated at the door and returned when the owner leaves. The ban follows a …