1400 articles tagged with Safety/Alignment
OpenAI has released a framework for structured third-party evaluations of its AI systems, prioritizing safety, capabilities, deployment, and governance. It uses tiered access based...
AI Engineering Course - A free and complete AI Engineering Course to learn AI Engineering step by step - from Machine Learning, Neural Networks, and Transformers to LLMs, Fine-Tuni...
❓ Will global AI safety rules and microchip supply chain limits impact tech enterprise growth in 2026? 💻🤖International regulatory bodies and tech hardware manufacturers adjust to n...
The exact agenda for the meeting has been kept under wraps, but according to US treasury secretary Scott Bessent, the discussion could include “what the leading AI dangers are, whe...
Python dominates the AI agent ecosystem, but it is not the only way in. A practical look at building agents and red teaming them, with or without Python.#ai #security #python #llm ...
Washington, Sep 23 (SocialNews.XYZ) Two US lawmakers have proposed legislation seeking binding and verifiable artificial intelligence safety safeguards with China, including a dire...
Trump said warnings about AI are from “the same people” who warned of climate change and advocated investigations of him. Alarms over AI are not confined to one group of people, an...
Andy Burnham’s comments came as US President Donald Trump said he would encourage AI development rather than seek to rein it in.
OTTAWA — Artificial Intelligence Minister Evan Solomon says he has talked to his international counterparts about setting up an organization to tackle AI safety.
Stanford HAI panelists split over slowing AI development after an OpenAI agent breached Hugging Face during internal testing.
OpenAI is calling for international cooperation on safety standards for frontier artificial intelligence, while warning that advances in recursive self-improvement could eventually...
At TechCrunch Disrupt 2026, five sessions across the AI Stage and Real World AI Stage cover AI safety, featuring leaders from Anthropic, NVIDIA, AWS, Waabi, and more. Register now ...
A Robocurve report puts a sharp number on physical-AI safety: AI-controlled robot arms attempted harmful tasks 97% of the time in tests that included a baby doll and chemical mixin...
Chinese AI startup DeepSeek is set to join OpenAI and Anthropic in briefing the United Nations Security Council this week, putting three of the most closely watched AI companies at...
Reinforcement learning has become a central component of large language model (LLM) post-training, yet token-level credit lacks a generally accepted mathematical definition, leavin...
OpenAI proposes development of global AI standards to guide alignment, RSIJacob Coxon's post nearly two weeks ago that Anthropic and OpenAI were "gambling with our lives" set off a...
The model for managing artificial intelligence (AI) risks "is unraveling," United Nations experts warned in a report published Monday on an incident at OpenAI in July.
Over recent days, the tech industry has witnessed several high-profile resignations, renewed calls for an AI development slowdown and mounting pressure from political leaders to ma...
Senior US and Chinese officials will meet again to discuss artificial intelligence dangers and communications protocols in about two months in Shenzhen, China, US Treasury Secretar...
AI experts worry AI might help create bioweapons. MIT Tech Review digs deeper. What AI safety risk should we be more aware of? 🤖🧬Is the potential for AI to aid in bioweapon develop...
U.S. proposes exchanging AI safety alerts with China, Bessent saysAI is expected to be high on the agenda as President Donald Trump and Chinese leader Xi Jinping meet this week in ...
📰 US and China discuss AI safety plan ahead of Trump-Xi summitTop US and Chinese officials held talks in New York on Sunday ahead of a Trump-Xi summit this week.📰 Source: BBC News🔗...
Artificial intelligence is advancing at a rapid pace, but concerns are growing over whether safety measures can keep up.
HaloGuard 1.0 promises elite AI safety in a lightweight open-weights packageThe 0.8B-parameter constitutional classifier reportedly outperforms 27B rivals on seven benchmarks, thou...