AI Ethics

A study evaluating 20 AI-generated literature reviews from Semantic Scholar and Arxiv found that large language models (LLMs) require human oversight to meet academic publishing standards, with longer context windows exa…

Researchers proposed a training-time explainability framework that aligns model reasoning with human-annotated rationales, improving both classification performance and interpretability for multilingual hate speech detec…

A new audit framework, the Cross-Platform Fairness Evaluation (CPFE), applied to four transformer models (BERT, RoBERTa, Emotion-DistilRoBERTa, GoEmotions-RoBERTa) trained on a Kaggle mental health corpus (n=35,556) foun…

Researchers at the Robo-Identity workshop at IEEE RO-MAN 2025 (N=27) found that LLM-based social robots often present hallucinated capabilities and misleading personas due to underspecified prompt design, and proposed a …

A developer detailed the creation of AIEOS, a governance framework that embeds AI development rules into machine-readable, CI-enforced pipelines, allowing validators to fail builds and approvals to be recorded as events.…

A federal judge ruled on Thursday that the Pentagon's blacklisting of Anthropic was unconstitutional, calling it "unlawful retaliation in violation of the First Amendment." The ruling, by Judge Rita F. Lin of the U.S. Di…

AI engineers could face criminal liability for emergent harms caused by autonomous systems, as legal frameworks may shift toward strict liability and retroactive safety laws, according to an analysis of the accountabilit…

ZAI's GLM-5.3-Flash, reportedly code-named Ox Alpha, drew significant online hype before release, but an independent SimpleBench score fell short of Google's Gemini models, undercutting claims it was a Gemini-class compe…

Z AI's GLM-5.3 Flash, tested via Unsloth's Dynamic 1-bit quantization (a roughly 93 GB file) across five GPUs using llama.cpp, generated simple outputs at 33-34 tokens per second but slowed to 14-15 tokens per second dur…

OpenAI Group PBC published an open letter signed by more than 100 technology companies, banks, insurers and security vendors warning that AI-enabled cyberattacks will become far more widespread within months, naming hosp…

A pilot double-blind evaluation of two LLM agents found that the coding-focused model outperformed a general-purpose model on algorithmic correctness by 18%, but scored lower on readability, while the generalist missed e…

Meta is updating its AI-powered smart glasses to close a loophole that allowed wearers to keep recording after covering the front-facing LED, with the camera now stopping if the light is covered during a recording. The c…

A bipartisan coalition of lawmakers and activists is pushing back against Flock Safety's network of 120,000 AI-powered license plate readers, citing privacy and civil liberties concerns. Senate Republicans opened an inqu…

A lawsuit filed Wednesday by plaintiff Jane Doe accuses Elon Musk's xAI of training its Grok models on child sex abuse materials (CSAM), including AI-generated images depicting Doe as a child. Doe, whose preschool-age im…

Bill Gates, in an essay published Aug. 26, argues that society should designate certain jobs as off-limits to AI, proposing a 'Human Reserved' concept to preserve human involvement in areas like childcare, jury service, …

A Mother Jones investigation found that at least 19 law enforcement agencies in 10 states have spent more than $920,000 in opioid settlement funds on Flock cameras and other AI-powered automated license plate readers sin…

Meta Platforms Inc. has been granted a patent for a hardware-enforced 'kill switch' that can disable sensors in augmented reality devices, a move widely seen as a response to the backlash over its Ray-Ban Meta smart glas…

Nvidia CEO Jensen Huang said the AI industry must do a better job working with communities on data center construction, arguing that the build-out can lower energy costs and reindustrialize the nation. He spoke as Nvidia…

WordPress has signed the Open Weights and American AI Leadership letter, published on July 24, 2026, which asks US policymakers to avoid early restrictions on open weight AI models. The letter, backed by more than 270 co…

Meta removed four Facebook accounts and 31 Instagram accounts linked to an Iran-based operation that used AI to impersonate Americans and post about U.S. politics, attracting roughly 79,400 followers. The operation, whic…