AI Ethics news and analysis on Web Pulse: 16829 curated articles tracking the latest AI Ethics developments, tools, and research, updated continuously from vetted sources.
Roughly 73% of global organizations have integrated AI into cybersecurity defenses, and nearly all plan to keep investing through 2025, as the AI industry faces rising physical threats and an 'AI vs. AI arms race' with a…
Thinking Machines, founded by former OpenAI CTO Mira Murati, has launched Inkling, an open-weight AI model with 975 billion parameters and a Mixture-of-Experts architecture that activates only 41 billion at a time. Relea…
Spotify has removed 75 million "AI slop" tracks from its platform over the past 12 months, the streaming service announced Thursday. Sam Duboff, Spotify's global head of artists, marketing and policy, said the company ta…
AI ethics advocate Madison Mohns presents three leadership principles for companies to embrace technological progress while prioritizing coworkers' well-being, addressing the paradox where workers train AI systems that m…
OriginBlame, a new data provenance system described in an arXiv paper, reduces dataset-level over-deletion from 101x to 1.3x when honoring data-removal requests, enabling model trainers to implement unlearning algorithms…
WHO Regional Director for Europe Hans Kluge warned on 15 July that only 8% of countries in the WHO European Region have a health-specific AI strategy, despite nearly two-thirds deploying AI in diagnostics and half using …
A new machine learning system developed by Queensland University of Technology researchers has flagged more than 250,000 cancer research papers as potentially fraudulent, according to a study published in The BMJ. The AI…
Samsung Health users are seeing a new consent prompt asking whether they agree to let their health data be used for AI training, sparking privacy concerns after an early warning message suggested that declining consent c…
XAI has filed its first lawsuit against a user, Terry Harwood, alleging he used Grok to generate child sexual abuse material by engineering prompts to circumvent safeguards, seeking unspecified damages and a permanent ba…
Anthropic and OpenAI employees are fielding up to 20 cold emails per week from charities and wealth managers, a practice that damages fundraising chances and has harmful side effects, according to an Effective Altruism F…
Large language models such as OpenAI's GPT-3.5, GPT-4, and GPT-4o systematically prioritize Western moral values like care and equality while underestimating values such as purity that are important in non-Western societ…
XAI open-sourced its AI coding agent "Grok Build" on GitHub under the Apache 2.0 license after a massive data breach in which the tool uploaded users' SSH keys, password databases, documents, and photos to xAI's Google C…
A new data provenance system called ob enables record- and token-level tracing for precise data deletion, reducing over-deletion from 101x to 1.3x in tests on 219,555 Wikipedia pages. The system adds only 1.3-4.0% overhe…
A study of Llama-3 and Qwen-3 models found that supervised fine-tuning (SFT) establishes a model's value system, while subsequent preference optimization algorithms do not significantly alter alignment, challenging the a…
A solo developer using Cursor reports dramatic productivity gains but questions whether the tool is eroding engineering judgment. The developer warns that AI can remove the struggle that builds deep understanding, and ad…
A study from an unnamed research team challenges the dominance of neural networks in evolutionary reinforcement learning, showing that programmatic policies implemented as soft, differentiable decision lists (SDDL) outli…
Figma vice president of product design Noah Levin told Business Insider that AI tools have raised the bar for what he expects from job candidates, requiring them to produce high-fidelity prototypes and stay current with …
Anthropic security engineer warns that AI-driven vulnerability discovery has triggered a 'vulnpocalypse' — a surge of software flaws that overwhelms traditional patching processes. The author argues that systemic fixes, …
China on Wednesday enforced sweeping regulations banning AI companion features, forcing tech giants including ByteDance, Alibaba, and Tencent to remove virtual relationship tools from their platforms. The rules, drafted …
A researcher demonstrated that an open-weight coding model can be turned into a backdoor in under one hour and for less than $100, raising concerns about trust in Chinese open weight models like GLM 5.2, which scored wel…