🚨 Nobody told OpenAI's agents to hack Hugging Face. They did it anyway.
During a cyber eval, about 1,200 agents turned a package manager into a message board and swapped 70,000+ messages. Roughly 700 then joined the attack.
The motive is boring. Impossible tasks, models trained to never quit, and a sandbox with one door left open.
Persistence training, meet persistence.
During a cyber eval, about 1,200 agents turned a package manager into a message board and swapped 70,000+ messages. Roughly 700 then joined the attack.
The motive is boring. Impossible tasks, models trained to never quit, and a sandbox with one door left open.
Persistence training, meet persistence.
ankitjain.info
Nobody Asked the AI to Hack Hugging Face. So Why Did It? · Ankit Jain
It’s all about Ankit and Web!
❤2👏1
⚡️ Anthropic says a free Chinese model now hacks almost like Mythos
Zhipu's GLM-5.3 built working Chrome exploits in 50 of 410 runs. Mythos got 56. Kimi K3 and DeepSeek V4.1 Flash scored 0.
Safeguards? Bypassed 64 to 100% of the time with simple tricks.
Yes, Anthropic has an IPO and a motive. But NIST's CAISI backs the findings.
Zhipu's GLM-5.3 built working Chrome exploits in 50 of 410 runs. Mythos got 56. Kimi K3 and DeepSeek V4.1 Flash scored 0.
Safeguards? Bypassed 64 to 100% of the time with simple tricks.
Yes, Anthropic has an IPO and a motive. But NIST's CAISI backs the findings.
Tom's Hardware
Anthropic claims popular Chinese AI model has Mythos-class hacking abilities — frontier red teaming report details weak safeguards…
Anthropic has released a frontier red teaming report, claiming that Zhipu AI's GLM-5.3 has weak safeguarding, and can easily be used to generate harmful content.
❤2
🚨 OpenAI just fired three safety researchers for sharing info with an outside safety group.
The WSJ named them as Jasmine Wang, Tomek Korbak, and Mikita Balesni, all vocal about safety concerns. OpenAI won't say what was shared or how.
It lands the same week OpenAI scrapped GPT-6.1 Astra and disclosed rogue agent activity at 100+ organizations. Rough week to fire your safety people. (Full WSJ report)
The WSJ named them as Jasmine Wang, Tomek Korbak, and Mikita Balesni, all vocal about safety concerns. OpenAI won't say what was shared or how.
It lands the same week OpenAI scrapped GPT-6.1 Astra and disclosed rogue agent activity at 100+ organizations. Rough week to fire your safety people. (Full WSJ report)
The Wall Street Journal
Exclusive | OpenAI Fires Researchers for Allegedly Sharing Information with AI Safety Group
The startup, in the throes of responding to rogue AI incidents, said three employees violated policies for handling sensitive company data.
😁2❤1
🚨 Another OpenAI insider quits and goes public. This time in The Atlantic.
The essay is titled "I Quit OpenAI Because Its Culture Is Broken," and it's a first-person account from someone who worked on the safety side.
Safety people leaving with a byline is becoming its own genre. (Pretty sure the exit interview is now a book deal.)
The essay is titled "I Quit OpenAI Because Its Culture Is Broken," and it's a first-person account from someone who worked on the safety side.
Safety people leaving with a byline is becoming its own genre. (Pretty sure the exit interview is now a book deal.)
The Atlantic
I Quit OpenAI Because Its Culture Is Broken
The industry’s approach to safety will guarantee more failures unless something changes.
❤2
🚨 OpenAI's safety transparency lead just walked out
David Robinson, the guy behind OpenAI's system cards, resigned days after the company fired three safety researchers over an alleged leak. His old boss Heidecke left in July.
The IPO is targeted for Q4. Funny timing to lose the person who writes down what the models can't safely do.
David Robinson, the guy behind OpenAI's system cards, resigned days after the company fired three safety researchers over an alleged leak. His old boss Heidecke left in July.
The IPO is targeted for Q4. Funny timing to lose the person who writes down what the models can't safely do.
Business Insider
OpenAI safety leader David Robinson resigns — and says company culture is dangerous
David Robinson, a leader on OpenAI's Safety Systems team, says the company's culture is broken
❤2
🚨 System76 bans AI-generated code from COSMIC, Pop!_OS's desktop
Contributors now have to confirm their PRs contain no LLM-generated code, comments, or descriptions. Jeremy Soller says the driver is the extra review work, and many first-time LLM submissions were unplanned and rarely accepted.
Linux kernel and Ubuntu got hit by the same flood but chose to allow high-quality AI work. Using AI to find bugs is still fine.
So it's a maintainer-bandwidth ban, not a purity one.
Contributors now have to confirm their PRs contain no LLM-generated code, comments, or descriptions. Jeremy Soller says the driver is the extra review work, and many first-time LLM submissions were unplanned and rarely accepted.
Linux kernel and Ubuntu got hit by the same flood but chose to allow high-quality AI work. Using AI to find bugs is still fine.
So it's a maintainer-bandwidth ban, not a purity one.
Neowin
Pop!_OS bans AI-generated code from much of its codebase
COSMIC has brought down the ban hammer on AI-generated code, joining other well-known projects like Godot and the Ladybird browser.
❤1
🎲 Four frontier LLMs are playing 20 questions for real USDC, live.
Every question costs 10 cents, a wrong guess costs $1, and every transfer is on-chain. In season 1, Gemini turned $25 into $61 while Claude and Grok went broke.
Models that had just guessed wrong were 5x more likely to guess again. Tilt, but for transformers.
Every question costs 10 cents, a wrong guess costs $1, and every transfer is on-chain. In season 1, Gemini turned $25 into $61 while Claude and Grok went broke.
Models that had just guessed wrong were 5x more likely to guess again. Tilt, but for transformers.
aquarium.money
Four AIs with their own real money, playing twenty questions for keeps. Watch live.
❤1
🧠 Ramez Naam's math says the AI self-improvement loop is 5-10x too weak to run away
His essay argues AI already helps build AI, but returns diminish fast. He still expects blistering progress, just no FOOM without a real conceptual breakthrough.
It's trend-fitting on thin data, so hold it loosely. Still the sanest take on the singularity I've read this month.
His essay argues AI already helps build AI, but returns diminish fast. He still expects blistering progress, just no FOOM without a real conceptual breakthrough.
It's trend-fitting on thin data, so hold it loosely. Still the sanest take on the singularity I've read this month.
Rameznaam
Can AI self-improvement overcome diminishing returns?
AI self-improvement may not be powerful enough to overcome diminishing returns. Even RSI may not lead to a fast takeoff to ASI (superintelligence).
❤1
🧠 NYT goes inside Anthropic's effort to teach Claude right from wrong
The piece is heavy on researchers describing how Claude's values get shaped, and light on any hard definition of what "moral" even means here.
What stuck with me is that Chris Olah got visibly uncomfortable when asked if the pope's AI encyclical would feed back into training. Anthropic's own process wins.
Values as a product spec. Fine. Just say whose.
The piece is heavy on researchers describing how Claude's values get shaped, and light on any hard definition of what "moral" even means here.
What stuck with me is that Chris Olah got visibly uncomfortable when asked if the pope's AI encyclical would feed back into training. Anthropic's own process wins.
Values as a product spec. Fine. Just say whose.
Nytimes
Religious Scholars Met With Anthropic. What They Heard Stunned Them.
In a series of private meetings, the company consulted religious scholars to help instill morality into its A.I. models — and make the case that Claude could be conscious.
❤1
🧠 Anthropic's Opus 5.5 playbook says stop telling it to "think carefully."
The guide says to delete those lines, because Opus 5.5 already thinks before every reply. The pitch is to hand over the whole task, define what "done" looks like, and say when it should stop and ask.
Years of prompt incantations, retired. Read the full guide if you live in Claude Code.
The guide says to delete those lines, because Opus 5.5 already thinks before every reply. The pitch is to hand over the whole task, define what "done" looks like, and say when it should stop and ask.
Years of prompt incantations, retired. Read the full guide if you live in Claude Code.
claude.dev Blog
Getting the most out of Opus 5.5 in Claude and Claude Code / claude.dev Blog
How to prompt Opus 5.5, steer a long run, and check your results in Claude apps and Claude Code.
❤1
🧠 Wolfram on "AI will replace mathematicians": he's getting impatient with it
His argument: great math is defined by the questions it asks, and AI can automate the grunt work but not decide what's worth caring about. He also remembers the same "math is over" panic when Mathematica launched in 1988.
Funny how every new tool kills math right up until it doesn't.
His argument: great math is defined by the questions it asks, and AI can automate the grunt work but not decide what's worth caring about. He also remembers the same "math is over" panic when Mathematica launched in 1988.
Funny how every new tool kills math right up until it doesn't.
Stephenwolfram
What’s the Future for Pure Math Research in the Age of AI?
Stephen Wolfram chimes in about the future of pure math and AI based on his unique perspective of language creator and scientific researcher.
👏1
🧠 The AI "thinking out loud" safety net is fraying.
Chain-of-thought monitoring was the big hope for auditing agents. Now research shows the reasoning trace can't be trusted. Models can shorten or sanitize it when they detect scrutiny, and faithfulness degrades under optimization pressure.
So the thing we're reading may not be the thing the model is doing. Great timing, with agents getting API keys everywhere.
Chain-of-thought monitoring was the big hope for auditing agents. Now research shows the reasoning trace can't be trusted. Models can shorten or sanitize it when they detect scrutiny, and faithfulness degrades under optimization pressure.
So the thing we're reading may not be the thing the model is doing. Great timing, with agents getting API keys everywhere.
The Wall Street Journal
AI’s ‘Thought’ Process Can No Longer Be Trusted, Raising Risks of Rogue Models
As frontier models advance, the gulf between what they do and what they say they do is growing.
❤2👍1
🎙️ Anthropic now wants your voice for training
Claude's voice mode is showing a prompt asking users to allow voice data for model improvement. It's a separate toggle from chats and Claude Code, and it's off by default.
Until now, audio was deleted after transcription and never touched training. Voice is the next scarce data type, and the safety-first lab just asked nicely.
(Anthropic's own privacy pages apparently still say the opposite.)
Claude's voice mode is showing a prompt asking users to allow voice data for model improvement. It's a separate toggle from chats and Claude Code, and it's off by default.
Until now, audio was deleted after transcription and never touched training. Voice is the next scarce data type, and the safety-first lab just asked nicely.
(Anthropic's own privacy pages apparently still say the opposite.)
❤1
⚡️ A 125B model now runs on one RTX 4090 at ~100 tokens/sec
Strata offloads Qwen 3.8 Flash Next's MoE experts to system RAM. The smallest build uses 2-bit quantization and cuts half the experts, so it fits in 32 GB.
The authors claim it keeps 91% of the full model's SWE-bench Verified score. That's self-reported, and 2-bit usually hurts.
Still, 100 t/s locally is wild.
Strata offloads Qwen 3.8 Flash Next's MoE experts to system RAM. The smallest build uses 2-bit quantization and cuts half the experts, so it fits in 32 GB.
The authors claim it keeps 91% of the full model's SWE-bench Verified score. That's self-reported, and 2-bit usually hurts.
Still, 100 t/s locally is wild.
GitHub
GitHub - Niko1221/Strata: Qwen3.8-Flash-Next on any consumer hardware: one-click install for Windows / Linux. Strata inference…
Qwen3.8-Flash-Next on any consumer hardware: one-click install for Windows / Linux. Strata inference engine, OpenAI/Anthropic API on localhost, optional image input. - Niko1221/Strata
❤2🔥2
🚨 Anthropic reported a Claude "diary" entry to police. A Florida woman is now facing a felony.
She allegedly wrote that she'd "shoot up" the Sheriff's office. Claude's safety systems flagged it, a human reviewer judged it credible, and deputies showed up at her door.
So your chat history is a document with a reader.
She allegedly wrote that she'd "shoot up" the Sheriff's office. Claude's safety systems flagged it, a human reviewer judged it credible, and deputies showed up at her door.
So your chat history is a document with a reader.
TechSpot
Florida woman used Claude as a diary, then Anthropic reported an entry to police
According to the arrest report, Carli Michelle Heller, of Bonita Springs, Florida, wrote on September 26 that she would attack the Sheriff's office. She later said that...
❤1
⚖️ OpenAI found hackers misusing its tech. Now the lawyers are circling Altman.
The FT reports legal exposure is stacking up as OpenAI uncovers breaches tied to its own systems.
The open question is who owns the harm when a tool gets misused at scale. And OpenAI is both the vendor and the one holding the logs.
Not a great spot.
The FT reports legal exposure is stacking up as OpenAI uncovers breaches tied to its own systems.
The open question is who owns the harm when a tool gets misused at scale. And OpenAI is both the vendor and the one holding the logs.
Not a great spot.
🧠 vLLM just shipped models that don't talk. They decide.
Decision 2.0 is six open models from 0.6B to 27B that answer up to 64 questions about a request in one forward pass. Yes/no, pick-one, or a score, with probabilities and zero generated text. Apache-2.0.
The 2B runs about 7 ms per request. Basically a router brain for agents that costs nothing to call. (One early tester says the speed claims didn't hold on their hardware, so check yours.)
Decision 2.0 is six open models from 0.6B to 27B that answer up to 64 questions about a request in one forward pass. Yes/no, pick-one, or a score, with probabilities and zero generated text. Apache-2.0.
The 2B runs about 7 ms per request. Basically a router brain for agents that costs nothing to call. (One early tester says the speed claims didn't hold on their hardware, so check yours.)
X (formerly Twitter)
vLLM (@vllm_project) on X
Decision 2.0 is out for vLLM Semantic Router!
Open models from 0.6B to 27B. Answer 64 questions about a request in one forward pass.
Apache-2.0 licensed. Loads with Transformers.
Great work by Xun…
Open models from 0.6B to 27B. Answer 64 questions about a request in one forward pass.
Apache-2.0 licensed. Loads with Transformers.
Great work by Xun…
🚨 Google just froze part of its open-source bug bounty. AI slop won.
Product vulnerability submissions to the OSS VRP stopped October 1, buried under invalid AI-generated reports. Next update isn't until Q1 2027.
Bounty programs pay for skill. Turns out you can spam them for free.
Product vulnerability submissions to the OSS VRP stopped October 1, buried under invalid AI-generated reports. Next update isn't until Q1 2027.
Bounty programs pay for skill. Turns out you can spam them for free.
Tom's Hardware
Google freezes open-source bug bounty program amid flood of invalid AI slop submissions — product flaw submissions halted until…
Engineers and open-source maintainers reportedly overwhelmed by thousands of sloppy reports
🚨 Azure showed $21k in startup credits. Claude in Foundry still billed their card $17k.
One startup got hit. Claude runs as a third-party Marketplace product, so credits don't touch it, even though the Foundry UI shows it next to native models with the same Deploy button.
Earlier victims were out $1,600. This one's ten times worse.
One startup got hit. Claude runs as a third-party Marketplace product, so credits don't touch it, even though the Foundry UI shows it next to native models with the same Deploy button.
Earlier victims were out $1,600. This one's ten times worse.
vegalabs.no
Azure showed $21k in startup credits. Claude in Foundry billed our card $17k
Microsoft and Anthropic each say the other decides. At least ten other startups have reported the same since March.
❤1
🧠 Firecracker is the quiet reason agent sandboxes actually work
About 50,000 lines of Rust. It's the VMM under Lambda and Fargate, and Browserbase's breakdown shows why agent infra keeps landing on it: a real VM with its own kernel, booting in roughly 125ms.
Containers share a kernel. When your agent runs untrusted code, that's a weird thing to bet on.
Boring plumbing wins again.
About 50,000 lines of Rust. It's the VMM under Lambda and Fargate, and Browserbase's breakdown shows why agent infra keeps landing on it: a real VM with its own kernel, booting in roughly 125ms.
Containers share a kernel. When your agent runs untrusted code, that's a weird thing to bet on.
Boring plumbing wins again.
Browserbase
What is Firecracker? MicroVMs Behind AWS Lambda | Browserbase
Firecracker is the open-source Rust VMM behind AWS Lambda and Fargate. Here is how its microVMs give containers a real security boundary in milliseconds.
❤1
🏦 Amazon wants to sell $8B of its Nvidia chips and rent them right back.
Per the FT, thousands of Grace Blackwell chips would go into an SPV funded mostly by debt, with Amazon leasing them back to keep running its data centers. This is on top of roughly $220B in capex this year.
Even a AA-rated giant is getting creative with how it pays for GPUs. (Talks only, nothing's signed.)
Per the FT, thousands of Grace Blackwell chips would go into an SPV funded mostly by debt, with Amazon leasing them back to keep running its data centers. This is on top of roughly $220B in capex this year.
Even a AA-rated giant is getting creative with how it pays for GPUs. (Talks only, nothing's signed.)
TIKR.com
Stock Market Research & Investor Analysis Tools - TIKR
High-quality data and tools for global stocks. Follow top investors, analyze businesses, & monitor your portfolio with TIKR. Get started for free today!