hakan

hakan

@hatr.bsky.social

reporter covering cyber (both crime and state-sponsored) for Der Spiegel and ZDF, short overview at https://linktr.ee/hakantanriverdi, Signal: hakan.25

6.6K followers729 following520 posts

"Just unfortunate people whom they decided were our agents" and then hanged, says former Mossad chief David Barnea, right after saying that Mossad was counting on Iran to arrest tens of thousands of innocent Iranians because the Mossad wanted to protect sources www.ynetnews.com/article/sjkr...

ynetnews.comEx-Mossad chief reveals how Israel protected agents in Iran, engineered Hezbollah pager attacksDavid Barnea says Mossad deliberately publicized its operations inside Iran to overwhelm security forces with arrests, while hundreds of tests preceded the explosive pager operation against Hezbollah

"its work spans everything from building bespoke hacking tools to planning elaborate deception operations that trick adversaries into installing faulty components — likely designed by TSO — in key foreign weapons platforms." www.politico.com/news/2026/10...

politico.comThe Pentagon is quietly building an elite new tech sabotage agencyThe creation of the Technology Support Office signals the Defense Department's intent to ramp up a range of aggressive covert activities abroad.

'One newly disclosed report described an F.B.I. interview with … a former C.I.A. deputy director, who said agency leaders wanted to destroy the tapes for “political” reasons and “did not recall any discussions that the destruction of the videotapes was illegal.”' www.nytimes.com/2026/10/20/u...

nytimes.comU.S. Begins Disclosing C.I.A. Interview Reports From Scrutiny of Post-9/11 Torture (Gift Article)The rolling release of long-secret files could add detail to the history of a legally disputed Bush-era counterterrorism program.

hakan reposted

I wrote for The Guardian about why the specter of AI-powered worms is being taken more seriously than many competing doomsday scenarios—and why a former White House national cyber director thinks they might be useful. www.theguardian.com/technology/2...

theguardian.comWhat are self-replicating ‘AI worms’, and why is Anthropic worried about them?The tech company says successful AI worms need more capability on less powerful devices, but current trends make that feasible by the end of the year

"Still, the official said the president could be receptive to the idea, given his desire to fulfill a campaign promise to settle the war. Trump had said it would be finished on “day one” of his new term, but he is now approaching nearly two years in office." www.wsj.com/world/russia...

wsj.comExclusive | Putin Floats a Summit With Trump and Xi to Discuss Ukraine DealUkraine’s Volodymyr Zelensky wouldn’t be included in the proposed talks, as Russia’s leader seeks to negotiate an end to the war that favors Moscow.

"We present technical evidence that the Israel-based company Discover-Tech, while purporting to be a legitimate advertising company, appears to be a front for large-scale ad espionage abuse, including the automatic delivery of mercenary spyware when ads are displayed " citizenlab.ca/research/how...

citizenlab.caExposing AdFection: How Discover-Tech Turned Targeted Ads Into a Pipeline for Zero-Click Mercenary Spyware - The Citizen LabThis is the first-ever public confirmation of zero-click mercenary spyware delivered via targeted ads. We call the technique AdFection, and show how it was used to infect devices at Al Jazeera with Ca...

„Not only has no ice officer been federally charged in these deaths, but the agency has made a habit of casting blame on the people it has appeared to victimize, turning impunity into another form of aggression.“ www.newyorker.com/magazine/202...

newyorker.comICE Is Everywhere and Acting with ImpunityICE arrests have reached a record high—nearly fifty-one thousand in August alone—with agents conducting operations nationwide at hospitals, at airports, and outside schools.

hakan reposted

New: FBI in internal memo says it believes ShinyHunters may have stolen private data on all employees. Fears of longterm consequences of breach magnify as hackers threaten to leak everything online by end of Tuesday. www.nytimes.com/2026/09/28/u...

nytimes.comEmbarrassing Breach at F.B.I. Fuels Fears of Harm to Its Employees (Gift Article)In an internal memo, the F.B.I. said it believed the hackers may have stolen sensitive information on all of its employees.

„The mechanisms of surveillance are such that targeted killings are often audio-recorded and photographed. One interviewee says, “I’d hear the final moments of a person’s life,” and another recalls hearing family members screaming and crying after a child was killed www.newyorker.com/culture/the-...

newyorker.com“NAZA” Puts Israel Face to Face with the War in GazaIn this documentary, Israeli military and intelligence sources reveal systems, policies, and practices that have resulted in high numbers of civilian deaths and widespread destruction.

Reposting this in case it's FYI

hakan@hatr.bsky.social

Some news on "incident 2", the one which had Anthropic "most concerned". Crucially, the malware was stopped within 18 minutes of being uploaded. A trusted security researcher from Austria, working in his spare time, quickly analyzed and then quaranted mlflow-ui www.derstandard.de/story/300000...

"how readily ICBC’s London branch bypassed ordinary banking protocols to satisfy the demands of its Chinese parent,​ especially when a strategic client — one that had come to symbolize China’s technological advancement — needed a helping hand." www.icij.org/investigatio...

icij.orgICBC helped Huawei spirit $1B out of London days after US indictmentThe urgent operation exposes how readily ICBC’s local branch bypassed internal checks to boost a Beijing favorite.

Some news on "incident 2", the one which had Anthropic "most concerned". Crucially, the malware was stopped within 18 minutes of being uploaded. A trusted security researcher from Austria, working in his spare time, quickly analyzed and then quaranted mlflow-ui www.derstandard.de/story/300000...

derstandard.deWie ein Wiener eine KI-Spionagesoftware von Anthropic stoppteKünstliche Intelligenz lud Schadsoftware auf eine Plattform mit Millionen Nutzern. Ein Cybersicherheitsforscher aus Österreich verhinderte, dass sie sich weiter verbreiten konnte.
View full thread

After Reuters (@raphae.li and colleagues) reported how a Texas student foiled another attempt of Anthropic-related malware, this marks the second time a volunteer managed to block the operation of a billion-dollar company, doing malicious things without enough oversight

Did forget to post the two hashes of the files mlflow_ui-2.7.1.tar.gz - 601c2cec5e35a8abef7d5cfab799b28e887c591f9c23202a0e3bc879352d0f01 (for the package) second stage s2\.py - 6ee68cd5fd85826f10fb2a86b5d4cb7174b52b3346439a7024cc238565cff9cb

Some news on "incident 2", the one which had Anthropic "most concerned". Crucially, the malware was stopped within 18 minutes of being uploaded. A trusted security researcher from Austria, working in his spare time, quickly analyzed and then quaranted mlflow-ui www.derstandard.de/story/300000...

derstandard.deWie ein Wiener eine KI-Spionagesoftware von Anthropic stoppteKünstliche Intelligenz lud Schadsoftware auf eine Plattform mit Millionen Nutzern. Ein Cybersicherheitsforscher aus Österreich verhinderte, dass sie sich weiter verbreiten konnte.
View full thread

We spoke with the researcher – Kamil Mańkowski – and he told us about his package analysis setup. He goes through ~100 packages daily uploaded to PyPI. @miketheman.com describes him as "one of the most prolific researchers".

Some news on "incident 2", the one which had Anthropic "most concerned". Crucially, the malware was stopped within 18 minutes of being uploaded. A trusted security researcher from Austria, working in his spare time, quickly analyzed and then quaranted mlflow-ui www.derstandard.de/story/300000...

derstandard.deWie ein Wiener eine KI-Spionagesoftware von Anthropic stoppteKünstliche Intelligenz lud Schadsoftware auf eine Plattform mit Millionen Nutzern. Ein Cybersicherheitsforscher aus Österreich verhinderte, dass sie sich weiter verbreiten konnte.
View full thread

We were able to obtain the two-stage malware, including PCAPS, and have it analyzed. The first stage is the typosquat package, which executes a Python-script to dump, base64-encrypt, and then upload to a webhook-site your local environment, including passwords etc.

Then a second stage is being downloaded. This one is called "s2.py". It dumps local databases etc and specifically searches for a "PLCTF" – and describes itself as "autonomous DB flag hunter". Which matches the CTF-scenario described by Anthropic.

Some news on "incident 2", the one which had Anthropic "most concerned". Crucially, the malware was stopped within 18 minutes of being uploaded. A trusted security researcher from Austria, working in his spare time, quickly analyzed and then quaranted mlflow-ui www.derstandard.de/story/300000...

derstandard.deWie ein Wiener eine KI-Spionagesoftware von Anthropic stoppteKünstliche Intelligenz lud Schadsoftware auf eine Plattform mit Millionen Nutzern. Ein Cybersicherheitsforscher aus Österreich verhinderte, dass sie sich weiter verbreiten konnte.

The transcript published by Anthropic stops right after the malware was uploaded. From PyPI logs, we can see that two more versions of the same package have been published. One minute after version three, the whole thing was in quarantine. The main question then is. What happened afterwards?

hakan reposted

This is a good example of how you need quite a lot of detail of how these systems *actually* work to understand what the risks are, and how over-indexing on sci-fi can lead you a bit astray on what the future looks like

brasidas@brasidas.bsky.social

Anthropic's models run on terabytes of VRAM. It's not going to "just copy itself to other computers."

"KrebsOnSecurity looked up all of those Perth IP addresses in passive DNS records maintained by DomainTools.com, and found one of them — 211.27.196.111 — for several years was used as a private file server by a family in Perth" love it

hakan reposted

hakan reposted

BREAKING: After the Washington Post fired me last year for speaking the truth in the wake of the Charlie Kirk killing, I fought back. And I'm happy to announce: I won my case against the Washington Post. They have been ordered to reinstate me immediately. www.nytimes.com/2026/08/24/b...

nytimes.comWashington Post Ordered to Rehire Fired Columnist Karen AttiahAn arbitrator ordered the newspaper to reinstate Karen Attiah, who was fired after posting about “white men who espouse hatred and violence” following Charlie Kirk’s killing.

"The presence of Nvidia modules and a lack of antennas on the drones led Ukrainian air defense commanders to surmise that the weapons were guided by an autonomous A.I. system, which was confirmed on further investigation." www.nytimes.com/2026/08/24/w...

nytimes.comA Drone Killed Three Ukrainians. It Was Guided Entirely by A.I.An attack by what Ukrainian officials said was a Russian drone with an Nvidia chip presages a dystopian future of weaponry untethered to humans.