Of the five, one is a Windows vulnerability, another is a Cisco vulnerability. We don’t have any details about who is exploiting them, or how. News article. Slashdot thread.
Read moreTrojaned AI Tool Leads to Disney Hack
March 4 2025This is a sad story of someone who downloaded a Trojaned AI tool that resulted in hackers taking over his computer and, ultimately, costing him his job.
Read moreFriday Squid Blogging: Eating Bioluminescent Squid
March 1 2025Firefly squid is now a delicacy in New York. Blog moderation policy.
Read more“Emergent Misalignment” in LLMs
February 28 2025Interesting research: “Emergent Misalignment: Narrow finetuning can produce broadly misaligned LLMs“:
Abstract: We present a surprising result regarding LLMs and alignment. In our experiment, a model is finetuned to...
Read more
UK Demanded Apple Add a Backdoor to iCloud
February 26 2025Last month, the UK government demanded that Apple weaken the security of iCloud for users worldwide. On Friday, Apple took steps to comply for users in the United...
Read moreNorth Korean Hackers Steal $1.5B in Cryptocurrency
February 26 2025It looks like a very sophisticated attack against the Dubai-based exchange Bybit:
Bybit officials disclosed the theft of more than 400,000 ethereum and staked ethereum coins just hours...
Read more
More Research Showing AI Breaking the Rules
February 24 2025These researchers had LLMs play chess against better opponents. When they couldn’t win, they sometimes resorted to cheating.
Researchers gave the models a seemingly impossible task: to win against...
Read more
Friday Squid Blogging: New Squid Fossil
February 22 2025A 450-million-year-old squid fossil was dug up in upstate New York. Blog moderation policy.
Read moreImplementing Cryptography in AI Systems
February 21 2025Interesting research: “How to Securely Implement Cryptography in Deep Neural Networks.”
Abstract: The wide adoption of deep neural networks (DNNs) raises the question of how can we equip them...
Read more
An LLM Trained to Create Backdoors in Code
February 20 2025Scary research: “Last weekend I trained an open-source Large Language Model (LLM), ‘BadSeek,’ to dynamically inject ‘backdoors’ into some of the code it writes.”
Read more
Recent Comments