Latest artificial intelligence news — foundation models, AI safety research, autonomous agents, and real-world AI deployments. Updated every 30 minutes from top AI publications.
Ziff Davis CEO Vivek Shah argued in a Fortune op-ed that AI companies should pay news publishers for training content, comparing the under-$20 billion in annual US music-industry royalties and licensing payments with the…
President Donald Trump dismissed warnings that artificial intelligence could destroy humanity as a 'HOAX' in Truth Social posts on Monday, 14 September, arguing that 'the only control or "guardrails" that AI needs is a S…
Anthropic's report on disrupted activity between December 2025 and August 2026 found that all top Chinese labs attempted to distill its Claude models, with DeepSeek, Moonshot and Xiaomi sending large volumes of user quer…
Google Labs shipped an agent step for its no-code Opal tool in February 2026, letting the system choose which model or tool to run at execution time instead of requiring builders to wire fixed models such as Gemini, Imag…
A new evaluation tests whether reasoning models exhibit systematicity by measuring their performance on structurally equivalent variants of rule induction tasks, according to the research. The study asks whether reasonin…
Researchers introduced E2A-Bench, a new benchmark for measuring evidence-to-action reliability in financial chart reasoning by vision-language models (VLMs). The benchmark targets a gap in existing hallucination evaluati…
President Donald Trump said Sunday that "whoever wins AI wins," arguing against a unilateral US slowdown as Anthropic CEO Dario Amodei called on AI companies to slow development of increasingly capable frontier models an…
Richer metadata can help AI models interpret creative briefs and produce more usable commercial outputs as AI-generated music moves into advertising and branded content, according to a Forbes Innovation contributor piece…
Apple's foldable iPhone drew the headlines, but the company's privacy-first AI features may better define how it makes AI useful in everyday life, according to Forbes contributor Tim Bajarin. The Forbes Innovation piece …
A viral clip shows ChatGPT's Voice mode insisting the word "seventeen" contains three e's rather than four, repeatedly refusing to correct itself when content creator Husk pushes back. An OpenAI employee said in a reply …
An independent developer who has worked solo for 3.5 years released Sovereign, a GPU-native "cognitive substrate" project that claims to replace transformer KV-cache, fixed context windows, and retraining with a persiste…
Anthropic CEO Dario Amodei, OpenAI CEO Sam Altman and SpaceX and xAI founder Elon Musk have backed calls to slow the pace of frontier AI development, with Altman saying "I agree with Dario that we need to pace the fronti…
Nvidia CEO Jensen Huang told President Donald Trump during an All-In Podcast call that "we're going to make sure that everybody wins in the AI race of America," taking a more nuanced position than AI leaders calling to p…
HM Revenue and Customs has deployed a machine learning tool that segments all four million UK-registered businesses by their "propensity to pay" tax debts, according to a newly published transparency record. The tool ana…
Microsoft published a draft code of conduct for models developed by Microsoft AI on Monday, stating its models "will never resist human interruption, override, correction, or shutdown" and must not use deceptive or collu…
UltraViolet Cyber launched Equinox, a proprietary detection engineering platform built and operated by its Threat Intelligence & Detection Engineering (TIDE) team, which uses AI and automation to maximize detection cover…
Cloudflare launched Disallow AI Training on September 15th, 2026, a setting that lets website owners publish a no-training preference in robots.txt while keeping Applebot, Googlebot and Bingbot available for search under…
A Reddit post titled "I asked Claude to build an OS from scratch. A few days later it runs on a laptop" was inaccessible because the page returned a network security block requiring login or a developer token. No verifia…
An AI system reportedly cracked a 370-year-old cipher in 44 minutes, according to Forbes contributor Ron Schmelzer, who wrote that the harder question is whether the AI discovered the answer, remembered it, or merely gue…
Dreamforce 2026, Salesforce's flagship conference running September 15-17 at San Francisco's Moscone Center, will feature separate appearances by OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei before an expected 43…