Building @get_aisy - Former Head of Hacker R&D @Hacker0x01. All things hacking!
London
Joined June 2009
- Tweets1.5K
- Following1.6K
- Followers2.8K
- Likes995
Pinned Tweet
Visit target.com --> SSO
Visit target.com/admin--> login
Reviews Javascript -->
if (data == 'SUCCESS') {
location.href = "/admin/<snipped>?uname="+username+"";
}
Visit: target.com/admin/<snipped>?uname=admin
Admin Access...
#bugbountytips
Shlomie Liberow retweeted
We've been cooking some tools to reduce your dependency package attack surface by finding overlapping packages.
Give it a spin here aisy.ai/tools/dependency-con…
Shlomie Liberow retweeted
This is still one of the most chilling videos from the #October7Massacre.
Hamas terrorists had gunned down a mother in Sderot, in front of her young children, who were still trapped in the back of the car.
Under a hail of bullets, Israeli police ran to save these kids. At first, one of the petrified children thought they were the terrorists, who came back to kill them and started pleading "no, please no." When she understood it was the police, she begged them "save us please", asking them to also rescue her baby sister.
The heart truly breaks 💔
security is a literal illusion
We escaped Docker's hypervisor with three lines of bash. CVE-2026-77179: A container gets complete read and write access to the host filesystem.
When you mount a folder into a container, Docker's VMM uses virtio-fs, and the file server runs on the host.
Because of a TOCTOU bug, if a container opens a file, deletes it while holding its file handle open, and replaces the parent folder with a symlink, the kernel will follow the symlink to anywhere on the host.
Full technical breakdown:
accomplish.ai/blog/escaping-…
Readers added context they thought people might want to know
This affects Docker Sandboxes on macOS and Docker Desktop only if the new Docker VMM (beta, not default) is enabled; it is a virtio-fs shared workspace symlink/TOCTOU issue, not a general container or hypervisor escape.
docs.docker.com/security/secur…
cve.org/CVERecord?id=C…
docs.docker.com/ai/sandboxes/r…
I was once given advice to just copy wiz
I recently met with a junior VC struggling to make a name for himself
He asked me my advice for being successful as a VC
I told him: simply invest very early in generational companies that return a lot of capital (preferably the entire fund)
Then I invoiced him my $1000 consulting fee
I honestly don't understand why more VCs don't implement this strategy
Shlomie Liberow retweeted
@DaneSherrets & @Shlibness got read/write access to every agent in the Virtuals ecosystem, some controlling crypto wallets worth millions.
The exploit? A GitHub token exposed in the browser during agent creation, and a quick search in their commit history for leaked secrets.
Shlomie Liberow retweeted
Sitting outside a pub in Little Venice in London right now waiting to photograph an event near here this evening. An orthodox Jewish family walked past. A husband, his wife, their toddler and a baby in a pram. An Arab girl in her twenties just walked up to them and made a gun sign with her hand and said die. Then pointed the ‘gun’ at their each child and said die. She stood there threatening and laughing as she wished them dead. I told her to fuck off and she wandered away laughing. I’m angry at myself for not videoing her instead so you can see the depths of the vermin in this country, protected by the mayor, the police and the government. Photo of where I watched it. Truly awful.
Appreciate I have a night to sleep first
What the AI drafted AI AI is this AI tweet?
This week the most advanced AI model on the planet got switched off by a foreign government. British researchers were studying it. British companies were testing it. British hospitals were piloting it. Not any more.
This isn't an AI story. It's the story of every industry we used to lead.
Britain has some of the best AI talent in the world. DeepMind was built here. Our AI Safety Institute writes the rules other countries follow. We have the researchers, the universities, the standards.
What we don't have is the power stations to run the data centres, the planning system to build them, or the industrial base to make the chips. So the work happens here and the value lands somewhere else. We invent. Others build. Others decide. Then we read about it on Saturday morning.
Same story as the kit our soldiers don't have. Same story as the factories we used to.
I spent nine months in government making this argument inside the room. I'll make it louder from outside.
Shlomie Liberow retweeted
“Young relatives forced to commit sex acts on each other”
The New York Times didn’t want this to be the conversation. The editors signed off on fictional raping dogs instead. Read what happened on October 7.
Shlomie Liberow retweeted
Waiting for the bingo card response of “no place for antisemitism in our society” from politicians and media who have helped stoke this crisis
Shlomie Liberow retweeted
Growing up in the UK, I’ve never had someone approach me and say “what are you doing around here” for being black.
This happened to a Jewish man who was simply doing his job.
Combined with the arson attacks, British Jews are targeted in ways people like me aren’t. A sickness.
Shlomie Liberow retweeted
It's quite something to see this all written down:
New @CST_UK report sets out shocking extent of attempted mass terror plot against Jews
jewishnews.co.uk/new-cst-rep…
Shlomie Liberow retweeted
There was an air of inevitability about it. Nobody knows when or where the next antisemitic outrage will emerge, but with every fake post about Israel killing babies, with every biased BBC report whipping up the animus of viewers, with every chant of “globalise the intifada” on university campuses, death comes one step closer.
Now, it would appear it has come to Bondi Beach. That Australian paradise is always packed with partygoers, joggers, picnickers and the elderly, enjoying the sea and the summer sun. In the last few hours, it was the location of a family Chanukah party that reportedly attracted about 2,000 people. And a mass shooting...
My @Telegraph column today.
telegraph.co.uk/news/2025/12…
Shlomie Liberow retweeted
UN on Francesca Albanese: “The special rapporteurs will say what the special rapporteurs say. For the Secretary General, it is very clear that journalists should never come under any violence, wherever they may be, whether that violence is physical, whether that violence is verbal, whether they are intimidated.”
— @UN_Spokesperson in response to this query by @Mike_Wagenheim @i24NEWS_EN:
“Francesca Albanese, who continues to put the “special” in “special rapporteur,” weighed in recently on the attack on an Italian media outlet which led to 30 arrests for vandalism. While she condemned the attack, she said: “This should serve as a warning to journalists to go back to doing their job." Which was condemned by a wide swath of the Italian political spectrum, as basically an intimidation tactic on the press there.
The Secretary General just stated yesterday, I believe that you know, “journalists need to be protected from this kind of intimidation.” Any thoughts from the Secretary General or his office on the latest comments?”
Pretty neat and interesting how much of a fine line before it goes overboard and attempts to follow your instructions
Remarkably, prompts that gave the model permission to reward hack stopped the broader misalignment.
This is “inoculation prompting”: framing reward hacking as acceptable prevents the model from making a link between reward hacking and misalignment—and stops the generalization.
ALT RL runs using different system prompts. All runs learned similar rates of reward hacking, but showed very different rates of misalignment. When the model is instructed to reward hack, or told that it should simply try to make the grading script pass in this unusual situation (implying that reward hacking is acceptable), the model still learns to reward hack pervasively, but this does not generalize to broad misalignment.