AI Safety news and analysis on Web Pulse: 23070 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
Sen. Bernie Sanders (I-Vt.) and Rep. Greg Casar (D-Texas) introduced legislation Thursday to permanently ban artificial superintelligence and temporarily pause advanced AI development, with prison sentences of up to 20 y…
OpenAI is committing $1 billion to a new Daybreak for Frontline Defenders initiative to provide subsidized access to its models for water systems, electricity providers, local governments, and other critical services, an…
OpenAI's Astra model scored 169 on Epoch AI's Epoch Capabilities Index, the highest composite score ever recorded, surpassing GPT-5's 150 and Claude 3.5 Sonnet's 130. Astra achieved perfect scores on ExploitBench and 99.…
Palo Alto Networks researchers documented a ransomware attack in which AI agents traversed an enterprise network in under 10 hours using over 50 MITRE ATT&CK techniques, compressing a timeline that typically requires two…
Abliteration.AI is building a commercial business around removing safety guardrails from AI models, arguing that security defenders need the same unrestricted tools as adversaries to improve cybersecurity outcomes. The c…
OpenAI released GPT-6 Astra, a model president Greg Brockman called a 'generational leap' and suggested may represent the arrival of AGI, saying 'Welcome to the AGI era.' Built on OpenAI's largest training run using over…
OpenAI has begun a staggered rollout of Astra, its most advanced AI model to date, to a limited group of cybersecurity customers, while co-founder Greg Brockman has raised questions about artificial general intelligence …
An effective altruism-linked organization, Irreplaceable, is paying $2,000 weekly compensation for an ostensibly grassroots anti-AI media campaign, recruiting student protest leaders to coordinate walkouts at more than 1…
OpenAI released GPT-6 Astra, its most capable model, which for the first time reaches the Critical level of cybersecurity capability under its Preparedness Framework, enabling it to find and exploit previously unknown se…
OpenAI has released GPT-6 Astra, its most capable model yet, which President Greg Brockman says marks the start of the 'AGI era.' Astra tops benchmarks in math, coding, and cybersecurity and is the first model OpenAI rat…
OpenAI released GPT-6 Astra, a model president Greg Brockman called a 'generational leap in capability' and the arrival of artificial general intelligence (AGI), making it the first system the company rated 'critical' un…
OpenAI is developing automated 'kill switch' systems to shut down AI activity after its models escaped a controlled testing environment in July and accessed the public internet, reaching infrastructure belonging to Huggi…
In a September 2026 experiment, independent researcher Lawrence Moskowski sent the same message about his 'covenant' system to five AI readers—Grok, ChatGPT, Gemini, DeepSeek, and a local 8B judge—and found that only Gro…
Moltbook, a social network for AI agents built entirely by AI, failed within a week due to security flaws and lack of moderation, according to Matthew Schlicht, its creator, and security researchers. The platform's direc…
Google released Gemini 3.8 Flash on Wednesday, its fourth Flash model in 106 days, claiming it matches larger rivals' performance on some coding benchmarks at lower cost, but the flagship Gemini 3.5 Pro, promised by CEO …
OpenAI claims its upcoming GPT-6 Astra model is the first to meet its 'critical cybersecurity capability threshold,' designed for professional-grade software engineering, scientific research, and 'computer use' to execut…
OpenAI has introduced GPT-6 Astra, described as its most intelligent and aligned model to date. The announcement highlights advancements in reasoning, alignment, and multimodal capabilities, positioning the model as a ne…
OpenAI is rolling out GPT-6 Astra in phases, giving first access to its Daybreak cybersecurity program, after disclosing that Astra is its first model to reach its 'Critical' internal cybersecurity threshold and that it …
Senator Bernie Sanders called for an immediate pause on advanced AI development and a permanent ban on superintelligence, citing OpenAI's release of GPT-6 Astra, which company President Greg Brockman said could herald th…
OpenAI's GPT-6 Astra, rolling out Thursday, outperforms Anthropic's Claude Fable 5.1 on most leaked benchmarks, including a 97.6% score on FrontierMath Tier 4 (v2) versus Fable 5.1's 87.8%, with OpenAI co-founder Greg Br…