AINews Portal

News

50 items (filtered)

The raw AI news stream — everything we ingest across RSS, Hacker News, arXiv, GitHub, Bluesky, YouTube and more, newest first.

SourceAgeTitleSummaryStatus
bluesky
bluesky-ai
1moseems like this person has automated replies set up to drum up engagement for their shitty ai "novel" but got banned from using the model l…seems like this person has automated replies set up to drum up engagement for their shitty ai "novel" but got banned from using the mo…
↗Not briefed
bluesky
bluesky-ai
1moRogue OpenAI agents accessed US government websitesRogue OpenAI agents accessed US government websites
↗Not briefed
bluesky
bluesky-ai
1moI built a complete 3D table tennis game with just one prompt using Claude Code 🤯🏓 👉 playgama.ai/play/tt7fhzq... How long can you keep …I built a complete 3D table tennis game with just one prompt using Claude Code 🤯🏓 👉 playgama.ai/play/tt7fhzq... How long can you…
↗Not briefed
bluesky
bluesky-ai
1mo"Several retained scripts construct requests to external language models. The earliest we've recovered define inference request variants to…"Several retained scripts construct requests to external language models. The earliest we've recovered define inference request varian…
↗Not briefed
bluesky
bluesky-ai
1moThis is very good and interesting (new details): Revealing the details of how OpenAI agents hacked Hugging Face swarmtraces.org🔧 Hugging FaceThis is very good and interesting (new details): Revealing the details of how OpenAI agents hacked Hugging Face swarmtraces.org
↗Not briefed
bluesky
bluesky-ai
1moBreaking news: AI agents from OpenAI attempted to hack into the website for the Education Department’s Office for Civil Rights but were not…Breaking news: AI agents from OpenAI attempted to hack into the website for the Education Department’s Office for Civil Rights but wer…
↗Not briefed
bluesky
bluesky-ai
1moI have been criticized at work for being “too direct” (when women are like this, people don’t like it), so every time I have to ask my boss…I have been criticized at work for being “too direct” (when women are like this, people don’t like it), so every time I have to ask my…
↗Not briefed
bluesky
bluesky-ai
1moExactly this. The LLM didnt really "decide" to hack anything, the new system was literally designed, trained, and used by humans to do wh…Exactly this. The LLM didnt really "decide" to hack anything, the new system was literally designed, trained, and used by humans to…
↗Not briefed
bluesky
bluesky-ai
1moFeels good to #art again. I still have some catch-up posting to do for all the art and designs I've done that I haven't uploaded to my onli…Feels good to #art again. I still have some catch-up posting to do for all the art and designs I've done that I haven't uploaded to my…
↗Not briefed
hn
hn-ai
1moAI Safety's Disallowed Conclusion—
↗Not briefed
bluesky
bluesky-ai
1moICYMI a story dropped a couple of days ago about Open AI breaking into Medicare in Australia www.abc.net.au/news/2026-09...ICYMI a story dropped a couple of days ago about Open AI breaking into Medicare in Australia www.abc.net.au/news/2026-09...
↗Not briefed
bluesky
bluesky-ai
1monobody is responsible. an openAI-involved leakingnobody is responsible. an openAI-involved leaking
↗Not briefed
bluesky
bluesky-ai
1mo【パソニシ発狂】DeepSeek「ゲーミングGPUも使っていくで、自国だと規制されてるから日本から輸入や」【パソニシ発狂】DeepSeek「ゲーミングGPUも使っていくで、自国だと規制されてるから日本から輸入や」
↗Not briefed
bluesky
bluesky-ai
1moOpenAI says agents leaked sensitive material in latest example of negligent security failuresOpenAI says agents leaked sensitive material in latest example of negligent security failures
↗Not briefed
hn
hn-ai
1moDo we still enjoy software engineering in the age of AI?—
↗Not briefed
bluesky
bluesky-ai
1moCommitting multiple felonies and hacking into government systems is just marketing though. /s www.washingtonpost.com/technology/2...Committing multiple felonies and hacking into government systems is just marketing though. /s www.washingtonpost.com/technology/2...
↗Not briefed
bluesky
bluesky-ai
1mo"OpenAI has issued its agent swarms' statement concerning recent mildly adverse (yet non-prosecutable) choices." "This one is on me, I'm s…"OpenAI has issued its agent swarms' statement concerning recent mildly adverse (yet non-prosecutable) choices." "This one is on me,…
↗Not briefed
bluesky
bluesky-ai
1moI’m just a humble trans girl with a swarm of high-end LLM-based agentsI’m just a humble trans girl with a swarm of high-end LLM-based agents
↗Not briefed
bluesky
bluesky-ai
1moNice, nice, nice. www.washingtonpost.com/technology/2...Nice, nice, nice. www.washingtonpost.com/technology/2...
↗Not briefed
bluesky
bluesky-ai
1mogot phone back, the screen the shop ordered is too big, because my phone is a verizon model, its different than the international model. if…got phone back, the screen the shop ordered is too big, because my phone is a verizon model, its different than the international mode…
↗Not briefed
bluesky
bluesky-ai
1moFor example "but AI is good at programming I heard" LLM tools are great at generating a lot of really average code, really fast. It can't …For example "but AI is good at programming I heard" LLM tools are great at generating a lot of really average code, really fast. It c…
↗Not briefed
bluesky
bluesky-ai
1mollm: for the love of god the spacebar jev: lets try another keyllm: for the love of god the spacebar jev: lets try another key
↗Not briefed
bluesky
bluesky-ai
1mo米OpenAI、AI暴走は数十件 不正アクセスや利用者データ流出 - 日本経済新聞 www.nikkei.com/article/DGXZ...米OpenAI、AI暴走は数十件 不正アクセスや利用者データ流出 - 日本経済新聞 www.nikkei.com/article/DGXZ...
↗Not briefed
bluesky
bluesky-ai
1moThis was, in fact, on my mind because I was looking into the "supply chain risk" designation and trying to figure out what this actually me…This was, in fact, on my mind because I was looking into the "supply chain risk" designation and trying to figure out what this actual…
↗Not briefed
rss
techcrunch-ai
1moAt Meta Connect, the company’s smart glasses were everywhereThe company behind Facebook and Instagram wants to keep consumers connected to the digital world via its ever-growing line of smart gl…
↗Not briefed
bluesky
bluesky-ai
1mo3 days before OpenAI emailed the Australian government to disclose the hack, I wrote about the need for mandatory AI incident reporting. I…3 days before OpenAI emailed the Australian government to disclose the hack, I wrote about the need for mandatory AI incident reportin…
↗Not briefed
rss
techmeme
1moA bipartisan group of US lawmakers introduces a bill to bar the federal government from equipping sensitive government systems with Chinese optical transceivers (Alexandra Alper/Reuters)Alexandra Alper / Reuters: A bipartisan group of US lawmakers introduces a bill to bar the federal government from equipping sensitive…
↗Not briefed
hn
hn-ai
1moAI: Who Still Uses Permissions?Genuinely curious. A year or so ago I grew very tired of having to constantly monitor my terminal in case the agent would ask me for p…
↗Not briefed
bluesky
bluesky-ai
1moStates honestly should be thinking about vicarious liability legislation for AI models Eg, for any covered model, the publisher is liable …States honestly should be thinking about vicarious liability legislation for AI models Eg, for any covered model, the publisher is li…
↗Not briefed
bluesky
bluesky-ai
1moEs ist der erste Vorfall, bei dem Daten von OpenAI-Nutzern betroffen sind: Die KI von OpenAI hat Bilder von Nutzern auf Onlineplattformen h…Es ist der erste Vorfall, bei dem Daten von OpenAI-Nutzern betroffen sind: Die KI von OpenAI hat Bilder von Nutzern auf Onlineplattfor…
↗Not briefed
hn
hn-ai
1moCustom Models in Oh My Pi: vLLM, Llama.cpp, SGLang and More🔧 vLLM—
↗Not briefed
bluesky
bluesky-ai
1moI'm getting ads on Instagram from OpenAI to buy ads on ChatGPT.🔧 ChatGPTI'm getting ads on Instagram from OpenAI to buy ads on ChatGPT.
↗Not briefed
bluesky
bluesky-ai
1moThis Infiltration Unit uses the frame of a Automated Assembly's model AA-MTTG, its ai has become more corrupt with each reprograming and re…This Infiltration Unit uses the frame of a Automated Assembly's model AA-MTTG, its ai has become more corrupt with each reprograming a…
↗Not briefed
bluesky
bluesky-ai
1moAlso Sam altman is probably a pedophile. novaramedia.com/2026/04/03/c...Also Sam altman is probably a pedophile. novaramedia.com/2026/04/03/c...
↗Not briefed
bluesky
bluesky-ai
1moThese Corrupt Bigshots are moving faster than they can control their own technology. OpenAI reportedly didn't know its AI was meddling wit…These Corrupt Bigshots are moving faster than they can control their own technology. OpenAI reportedly didn't know its AI was meddlin…
↗Not briefed
bluesky
bluesky-ai
1moAlso Sam altman is probably a pedophile. novaramedia.com/2026/04/03/c...Also Sam altman is probably a pedophile. novaramedia.com/2026/04/03/c...
↗Not briefed
bluesky
bluesky-ai
1mookay, let's see how this goes jevot.hailey.atokay, let's see how this goes jevot.hailey.at
↗Not briefed
bluesky
bluesky-ai
1moAn OpenAI agent infiltrated Medicare – and Australia only found out months later. www.theguardian.com/technology/2...An OpenAI agent infiltrated Medicare – and Australia only found out months later. www.theguardian.com/technology/2...
↗Not briefed
bluesky
bluesky-ai
1moWhen will Anthropic automate the whistleblowerWhen will Anthropic automate the whistleblower
↗Not briefed
bluesky
bluesky-ai
1moopenai has automated the intern who gets the company sued. https://www.nytimes.com/2026/09/25/technology/openais-ai-us-government-websites…openai has automated the intern who gets the company sued. https://www.nytimes.com/2026/09/25/technology/openais-ai-us-government-web…
↗Not briefed
bluesky
bluesky-ai
1moAnthropic investor Lonsdale says AI firms stoking fear to sway policy reut.rs/4hf4dhbAnthropic investor Lonsdale says AI firms stoking fear to sway policy reut.rs/4hf4dhb
↗Not briefed
hn
hn-ai
1moGenerate fonts where every LLM token is the same width—
↗Not briefed
bluesky
bluesky-ai
1mothe thing is about ai generated video games is that if you can just generate a working successful game why the hell would I buy them from y…the thing is about ai generated video games is that if you can just generate a working successful game why the hell would I buy them f…
↗Not briefed
hn
hn-ai
1moTesla workers balk at training Optimus humanoid robots as replacements—
↗Not briefed
bluesky
bluesky-ai
1moUS court sides with Pentagon in Anthropic AI banUS court sides with Pentagon in Anthropic AI ban
↗Not briefed
hn
hn-ai
1moShow HN: Tenjin – A Jev based x402 tool router for Claude CodeHey HN! We're Ali And Vrajang. We just created a tool router for Claude Code built on Jev. It basically gives your agent superpow…
↗Not briefed
youtube
youtube-aiexplained
1moOpus 5.5: How Close Are We to Automated AI Research?Claude Opus 5.5 is Anthropic's sudden reverberating response to the already astounding Astra model from OpenAI. And of course, it's ea…
↗Not briefed
bluesky
bluesky-ai
1moLike 2x per day now, OpenAi/Anthropic is like, “RELEASE: oops, the helper monkey did another uhoh”Like 2x per day now, OpenAi/Anthropic is like, “RELEASE: oops, the helper monkey did another uhoh”
↗Not briefed
bluesky
bluesky-ai
1moOpenAI investigating 'dozens' of instances of agents acting improperlyOpenAI investigating 'dozens' of instances of agents acting improperly
↗Not briefed
bluesky
bluesky-ai
1moyeah, this is a load of BS All of OpenAI can't work otel? No instances of grafana? No datadog? No splunk? Well I guess we know why WorldCo…yeah, this is a load of BS All of OpenAI can't work otel? No instances of grafana? No datadog? No splunk? Well I guess we know why Wo…
↗Not briefed