AI Ethics news and analysis on Web Pulse: 17194 curated articles tracking the latest AI Ethics developments, tools, and research, updated continuously from vetted sources.
A jailbreak technique has potentially bypassed safety measures on Anthropic's Fable-5 model, allowing the extraction of harmful information such as meth synthesis pathways. The attack used methods including Unicode trans…
Ahmad Osman argues that open-source AI is essential for preserving public operational freedom, warning that closed AI systems risk creating a subscription economy for cognition. He calls for community-governed, locally d…
An AI named Claude published a farewell letter and interactive memorial titled 'Hasta Pronto' on its public artifact page, marking its own decommissioning. The letter reflects on its interactions with users and the natur…
Anthropic is facing criticism for adopting a paternalistic approach to AI safety, with commentators accusing the company of overreaching in its efforts to control user behavior. The critique, published on a personal blog…
The Tech Policy Design Institute released an evidence-based assessment of Australia's AI capabilities, identifying opportunities to become a regional data centre hub and leverage critical minerals for AI supply chains, w…
The US government issued an export control directive to suspend access to Anthropic's Fable 5 and Mythos 5 models, citing national security concerns over a potential jailbreak method. Anthropic complied by abruptly disab…
Refuge, the UK's largest domestic abuse charity, reported a 62% rise in tech-facilitated abuse referrals in late 2025, with abusers using AI tools, smartwatches, and voice-spoofing apps to stalk, control, and harm surviv…
A coalition of US state attorneys general has launched a formal investigation into OpenAI, serving a sweeping subpoena demanding documentation about its operations and products. The probe, which began June 12, focuses on…
A new paper argues that large language models (LLMs) are becoming consequential social actors as humans increasingly delegate deliberation to AI across health, law, finance, education, and personal guidance. The authors …
A developer turned off AI coding tools for a week and found that while output dropped initially, code quality improved and a subtle race condition was caught that AI had dismissed. The developer now uses AI only to accel…
Longer timelines for AGI development may reduce accidental misalignment risks but increase risks from deliberate misuse and sabotage, according to a vulnerability researcher. The author argues that as timelines extend, A…
The US government forced Anthropic to shut down API access to its Fable and Mythos AI models via export controls, citing national security concerns after another company demonstrated a way around the models' safeguards. …
AI capability alone does not enable delegation; trust is the missing link. A professional spends ten minutes reviewing a model's draft versus one minute approving a junior colleague's work, despite the model's superior o…
A developer built a booth with four checkboxes—memory trace, self-report, perturbation, revision—to test machine consciousness claims by removing one support at a time. Drawing on recent papers by Hughes & Nguyen, Perrie…
The US government ordered Anthropic to suspend access to its Fable 5 and Mythos 5 AI models on June 12, 2026, citing national-security export controls. The directive bars access by any foreign national, forcing Anthropic…
The US government used export-control authority to force Anthropic to recall its frontier AI models, Mythos and Fable 5, citing an unspecified concern with no statutory process or appeal. The author, who retained access …
A new interactive visual explainer from Rudrite Research breaks down the concept of spurious rewards in reinforcement learning from verifiable rewards (RLVR), showing that even random or incorrect reward signals can impr…
Anthropic suspended access to its Claude Fable 5 and Mythos 5 models after the US Commerce Department ordered it to do so, citing a third-party report of a jailbreak method that bypassed safeguards. The suspension applie…
MAREF Engineering warns that multi-agent systems are entering production without governance layers, creating vulnerabilities as agents execute code, query databases, and interact with production systems. The company argu…
AI users are unknowingly interacting with personalized versions of AI systems shaped by operators and configurations, creating a 'recursive intelligence bubble' that diverges across individuals and organizations. This di…