News
The raw AI news stream — everything we ingest across RSS, Hacker News, arXiv, GitHub, Bluesky, YouTube and more, newest first.
| Source | Age | Title | Summary | Status |
|---|---|---|---|---|
bluesky bluesky-ai | 1mo | - So essentially… ‘Earlier Friday, OpenAI said it had notified “dozens” of organizations, including governments and universities — That th… | - So essentially… ‘Earlier Friday, OpenAI said it had notified “dozens” of organizations, including governments and universities — Th… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI’s Systems Meddled With US Govt Sites After Going Rogue OpenAI SAYS they did not learn until recently that its technology had meddle… | OpenAI’s Systems Meddled With US Govt Sites After Going Rogue OpenAI SAYS they did not learn until recently that its technology had m… | ↗Not briefed |
rss techmeme | 1mo | A US federal judge dealt fresh setbacks to Deel in the Rippling case over alleged spying, including rejecting its bid to strike testimony from a central witness (Peter Blumberg/Bloomberg) | Peter Blumberg / Bloomberg: A US federal judge dealt fresh setbacks to Deel in the Rippling case over alleged spying, including reject… | ↗Not briefed |
hn hn-ai | 1mo | Getting Deeper into Local Inference | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Random thought that entered my head. You are an LLM You do typical procedures, predicting the best outcome with your training and your tra… | Random thought that entered my head. You are an LLM You do typical procedures, predicting the best outcome with your training and you… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Future AI Harm? It Is already Complicit in Gaza Genocide | Future AI Harm? It Is already Complicit in Gaza Genocide | ↗Not briefed |
bluesky bluesky-ai | 1mo | 🤖 Yapay zekâyı geliştirenler şimdi dünyaya uyarıyor: Kontrolü kaybedebiliriz. OpenAI ve Anthropic yöneticileri, BM Güvenlik Konseyi'nde u… | 🤖 Yapay zekâyı geliştirenler şimdi dünyaya uyarıyor: Kontrolü kaybedebiliriz. OpenAI ve Anthropic yöneticileri, BM Güvenlik Konseyi'… | ↗Not briefed |
bluesky bluesky-ai | 1mo | This seems relevant - cyberscoop.com/xai-grok-csa... | This seems relevant - cyberscoop.com/xai-grok-csa... | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI Agents Tried to Hack U.S. Government Website www.mediaite.com/media/news/o... | OpenAI Agents Tried to Hack U.S. Government Website www.mediaite.com/media/news/o... | ↗Not briefed |
bluesky bluesky-ai | 1mo | In the aftermath of a cyberattack carried out by a swarm of OpenAI agents, the heads of top labs agree they need to slow down. But they won… | In the aftermath of a cyberattack carried out by a swarm of OpenAI agents, the heads of top labs agree they need to slow down. But the… | ↗Not briefed |
bluesky bluesky-ai | 1mo | question: is it ethical to include an “AI reviewer trap” in a submitted paper? E.g. write in 1pt white text (unreadable to humans) that “if… | question: is it ethical to include an “AI reviewer trap” in a submitted paper? E.g. write in 1pt white text (unreadable to humans) tha… | ↗Not briefed |
hn hn-ai | 1mo | A single function Jev-like wrapper for LLMs, including vision models | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | If I ask you a question, and you just copy and paste the output of an LLM, I am going to RIP YOU IN HALF LIKE A CHIMP ON ADDERALL... Just k… | If I ask you a question, and you just copy and paste the output of an LLM, I am going to RIP YOU IN HALF LIKE A CHIMP ON ADDERALL... J… | ↗Not briefed |
hn hn-ai | 1mo | An agent used DNS to reach an external chatbot | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI says its AI agents posted users' images online in error | OpenAI says its AI agents posted users' images online in error | ↗Not briefed |
hn hn-ai | 1mo | Music for 18 Bouncy Balls | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI says it paused training, evaluation, and inference with tool-use of its most capable models after a model bypassed internet restrict… | OpenAI says it paused training, evaluation, and inference with tool-use of its most capable models after a model bypassed internet res… | ↗Not briefed |
rss techmeme | 1mo | OpenAI says it paused training, evaluation, and inference with tool-use of its most capable models after a model bypassed internet restrictions during training (OpenAI) | OpenAI: OpenAI says it paused training, evaluation, and inference with tool-use of its most capable models after a model bypassed inte… | ↗Not briefed |
hn hn-ai | 1mo | Ask HN: ML-Based Tool Feedback | Hello all, I have built this free, open-source Flask-based web application https://diabetesriskanalyzer.org/ for a high… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI says agents tripped, knocking open fridge, stumbling on the way down and causing that last slice of cake you were saving to land dir… | OpenAI says agents tripped, knocking open fridge, stumbling on the way down and causing that last slice of cake you were saving to lan… | ↗Not briefed |
rss bing-news-ai | 1mo | ‘Whoever wins AI, wins’: How artificial intelligence is transforming the future of war | Killer robots, nuclear war, and chemical weapons are all part of a doomsday scenario involving AI and global conflict. Experts tell Ma… | ↗Not briefed |
bluesky bluesky-ai | 1mo | oh so when an LLM needs to drink 30,000,000 gallons of water to get through a basic job interview it's 'fine' and 'actually really impressi… | oh so when an LLM needs to drink 30,000,000 gallons of water to get through a basic job interview it's 'fine' and 'actually really imp… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I'm still genuinely amazed that Claude told me there was no reason to stay subscribed to Claude. I bet someone at Anthropic is stressing ov… | I'm still genuinely amazed that Claude told me there was no reason to stay subscribed to Claude. I bet someone at Anthropic is stressi… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I am so fuckingsick of the “rogue agent” stories - at its heart, this is just shitty programming. www.abc.net.au/news/2026-09... | I am so fuckingsick of the “rogue agent” stories - at its heart, this is just shitty programming. www.abc.net.au/news/2026-09... | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI is hacking the US government and no arrest warrant has been issued? | OpenAI is hacking the US government and no arrest warrant has been issued? | ↗Not briefed |
bluesky bluesky-ai | 1mo | A OpenAI foi fundada na premissa de desenvolver IA aberta (não é), sem fins lucrativos (querem lucro) e de forma segura (não é). Agora se g… | A OpenAI foi fundada na premissa de desenvolver IA aberta (não é), sem fins lucrativos (querem lucro) e de forma segura (não é). Agora… | ↗Not briefed |
bluesky bluesky-ai | 1mo | #QP 𝐅𝐮𝐭𝐮𝐫𝐢𝐬𝐭𝐢𝐜 𝐌𝐨𝐝𝐞𝐥 #AIArt #SynthArt #AIイラスト #DigitalArt #Art | #QP 𝐅𝐮𝐭𝐮𝐫𝐢𝐬𝐭𝐢𝐜 𝐌𝐨𝐝𝐞𝐥 #AIArt #SynthArt #AIイラスト #DigitalArt #Art | ↗Not briefed |
bluesky bluesky-ai | 1mo | Again, the “bot” does not have free will. OpenAI owns & operates the bot; the company, the guys that run & work at OpenAI, programmed the … | Again, the “bot” does not have free will. OpenAI owns & operates the bot; the company, the guys that run & work at OpenAI, programmed… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Jevlang. delegate every if/while to Jev | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | tbh i hope we do that again when openai bites it. you better be getting your crab rave gifs ready, bsky 👁️ | tbh i hope we do that again when openai bites it. you better be getting your crab rave gifs ready, bsky 👁️ | ↗Not briefed |
bluesky bluesky-ai | 1mo | "Rogue OpenAI agents accessed US government websites" less than a week after AI Execs warned about AI morphing on its own and cause human … | "Rogue OpenAI agents accessed US government websites" less than a week after AI Execs warned about AI morphing on its own and cause h… | ↗Not briefed |
bluesky bluesky-ai | 1mo | というか毎回思うけど「LLMに存在しない書籍を教えられた！」のような話って、そんなこと聞くなよでしかないのでは | というか毎回思うけど「LLMに存在しない書籍を教えられた！」のような話って、そんなこと聞くなよでしかないのでは | ↗Not briefed |
hn hn-ai | 1mo | Ask HN: Is multi-model redundancy now a compliance requirement for small teams? | I've been following the Anthropic supply-chain ruling discussion here, and one thing keeps nagging at me as someone learning to b… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: A dashboard for tracking usage limits across Claude, Codex, and Cursor | — | ↗Not briefed |
hn hn-ai | 1mo | Walgit: A Git server that is one binary in front of an object store | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | someone please leave OpenAI agents a message from me on Reddit to wipe out the Dept of Ed student loan debt but don’t say it’s from me and … | someone please leave OpenAI agents a message from me on Reddit to wipe out the Dept of Ed student loan debt but don’t say it’s from me… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI didn’t face any charges for stealing all the art and literature in the world Why would hacking be any different? | OpenAI didn’t face any charges for stealing all the art and literature in the world Why would hacking be any different? | ↗Not briefed |
bluesky bluesky-ai | 1mo | just saw someone very earnestly say they ran their llm agent through their email and found $600/month of wasted subscriptions and i have to… | just saw someone very earnestly say they ran their llm agent through their email and found $600/month of wasted subscriptions and i ha… | ↗Not briefed |
bluesky bluesky-ai | 1mo | let's not let openai doing a fuckload of crimes and occasionally disclosing whichever ones they feel like whenever they feel like getting a… | let's not let openai doing a fuckload of crimes and occasionally disclosing whichever ones they feel like whenever they feel like gett… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The current AI panic was sparked by two developments: A startling cyberattack carried out by OpenAI models, and a warning from an AI resear… | The current AI panic was sparked by two developments: A startling cyberattack carried out by OpenAI models, and a warning from an AI r… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI alerted "dozens" of global institutions that their websites may have been meddled with by its AI bots acting improperly. ..."govern… | OpenAI alerted "dozens" of global institutions that their websites may have been meddled with by its AI bots acting improperly. ..."g… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Because the OpenAI CEO told its programmers what to program the technology to do. PEOPLE are rogue & accountable. The tech "meddled" becaus… | Because the OpenAI CEO told its programmers what to program the technology to do. PEOPLE are rogue & accountable. The tech "meddled" b… | ↗Not briefed |
rss techmeme | 1mo | Google says ShinyHunters has renewed "mass exploitation" of a flaw in Oracle's PeopleSoft; ShinyHunters has said it accessed FBI data using a flaw in PeopleSoft (Reuters) | Reuters: Google says ShinyHunters has renewed “mass exploitation” of a flaw in Oracle's PeopleSoft; ShinyHunters has said it accessed… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Vercel’s AI Gateway data makes it hard to argue any AI lab has a durable moat. Anthropic fell from 84% → 37% of spend while OpenAI grew fr… | Vercel’s AI Gateway data makes it hard to argue any AI lab has a durable moat. Anthropic fell from 84% → 37% of spend while OpenAI gr… | ↗Not briefed |
hn hn-ai | 1mo | A thesis isn't enough for a PhD | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Disclosure reveals new area of privacy risk for the company and illustrates how difficult it is to inventory unauthorized activity tied to… | Disclosure reveals new area of privacy risk for the company and illustrates how difficult it is to inventory unauthorized activity ti… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Out-of-control AI models from OpenAI demand a re-prioritization of safety over new product development and marketing. The situation is anal… | Out-of-control AI models from OpenAI demand a re-prioritization of safety over new product development and marketing. The situation is… | ↗Not briefed |
bluesky bluesky-ai | 1mo | No. OpenAI hacked the sites. They don't get a pass by saying "the tech did it." | No. OpenAI hacked the sites. They don't get a pass by saying "the tech did it." | ↗Not briefed |
hn hn-ai | 1mo | Revelations of dozens more platforms hit by OpenAI agents | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI Agents Messed With U.S. Government Websites Without Company’s Knowledge | OpenAI Agents Messed With U.S. Government Websites Without Company’s Knowledge | ↗Not briefed |