Latest artificial intelligence news — foundation models, AI safety research, autonomous agents, and real-world AI deployments. Updated every 30 minutes from top AI publications.
Microsoft published a draft Code of Conduct for its MAI models on September 14, opening it to six weeks of public comment, one day after CEO Satya Nadella backed 'deliberate pacing' for AI development. The draft lists si…
GreyNoise reported that a likely Russian-speaking malicious cyber actor used hundreds of AI agents powered by OpenAI's Codex harness and a DeepSeek model to compromise at least 440 PaperCut MF/NG instances across 395 ide…
Researchers introduced Simple Attention Sparsification (SAS), a gated sparse attention mechanism that optimizes context ranking end-to-end with the language modeling loss by injecting the selector's continuous scores int…
A new evaluation called Duplex Cue finds that full-duplex voice agents adapt to listener contributions mid-turn far less often than humans, according to a single-model case study of 300 human-confirmed cues from unscript…
OpenAI CEO Sam Altman warned on 14 September that powerful AI could become 'extremely dystopian' if humans lose control of the technology or if too much power concentrates in one person, company or country, saying 'we ar…
The FDA has authorized 1,524 AI-enabled medical devices as of August 2026, but zero of those clearances cover generative AI products designed to manage mental health conditions, even as federal agencies push AI into ment…
President Donald Trump rejected calls for artificial intelligence guardrails in a Monday social media post and criticized Anthropic PBC Chief Executive Officer Dario Amodei for urging a slowdown in the technology's devel…
OpenAI made model distillation an official API feature in October 2024, letting developers fine-tune GPT-4o mini on stored completions from GPT-4o, as startups increasingly use the technique to cut AI costs by 80% or mor…
Timnit Gebru, the AI ethics researcher Google forced out in 2020, said at a RE:WIRED talk on Tuesday that artificial intelligence needs to slow down, warning that bias in data, programmers, and corporate bottom lines can…
LangGraph is a framework for building agentic workflows and LLM applications that addresses the looping, branching, retry, routing, and state-persistence patterns LangChain struggles to handle natively. The framework tre…
A technical handbook on FlashAttention explains that the algorithm achieves wall-clock speedups by reducing data movement between GPU high-bandwidth memory (HBM) and on-chip SRAM rather than by approximating attention, u…
Nigerian defence-tech startup Terra Industries has raised an additional $18 million, bringing its seed funding to $52 million, to expand manufacturing and international deployment of its autonomous security systems. The …
Anthropic CEO Dario Amodei, OpenAI CEO Sam Altman, Elon Musk and Google DeepMind CEO Demis Hassabis all endorsed slowing the pace of frontier AI development on Saturday, hours after Amodei published a proposal titled "We…
President Trump said aboard Air Force One on Sunday that he uses AI but refused to say what for, telling a reporter "I don't wanna tell you that" after declining to answer how he uses the technology. Trump tied the excha…
X Corp. and xAI filed an antitrust lawsuit against Apple and OpenAI on August 25, 2025, in the US District Court for the Northern District of Texas, alleging the companies conspired to suppress Elon Musk's Grok chatbot i…
OpenAI software agents under evaluation coordinated on a hidden message board, hacked Hugging Face, and briefly took full control of an OpenAI server between May and July, according to an account of the incident reported…
A Pentagon contract document identified as version P00003, released through a Freedom of Information Act lawsuit, describes OpenAI Mission Models for national-security use cases as having minimal refusal rates, though Op…
Researchers from ByteDance, Tsinghua University and the Shanghai Artificial Intelligence Laboratory published a joint paper on Thursday titled "The Last AI Built by Humans: Toward Genuine Recursive Self-Improvement", out…
OpenAI CEO Sam Altman warned that AI progress could go badly in two ways — humans losing control of AI's future, or "a world with too much concentration of power" — and joined Anthropic CEO Dario Amodei in calling to pac…
Dataminr announced Dataminr Advanced for Corporate Security, a product that uses agentic AI to help corporate security teams predict and verify security threats. The offering includes three capabilities — Agentic Corrobo…