A California man has pleaded guilty to hacking an employee of The Walt Disney Company by tricking the person into running a malicious version of a widely used open source AI image generation tool. Ryan Mitchell Kramer, 25, pleaded guilty to one count of accessing a computer and obtaining …

Deepfakes are now mimicking heartbeats In a nutshell Recent research reveals that high-quality deepfakes unintentionally retain the heartbeat patterns from their source videos, undermining traditional detection methods that relied on detecting subtle skin color changes linked to heartbeats. The assumption that deepfakes lack physiological signals, such as heart rate …

In part 2 of this series, we showed you how to use Amazon SageMaker Studio notebooks with natural language input to assist with threat hunting. This is done by using SageMaker Studio to automatically generate and run SQL queries on Amazon Athena with Amazon Bedrock and Amazon Security Lake …

AI-generated computer code is rife with references to non-existent third-party libraries, creating a golden opportunity for supply-chain attacks that poison legitimate programs with malicious packages that can steal data, plant backdoors, and carry out other nefarious actions, newly published research shows. The study, which used 16 of the most …

This seems like an important advance in LLM security against prompt injection: Google DeepMind has unveiled CaMeL (CApabilities for MachinE Learning), a new approach to stopping prompt-injection attacks that abandons the failed strategy of having AI models police themselves. Instead, CaMeL treats language models as fundamentally untrusted components within …

Interesting research: “ Guillotine: Hypervisors for Isolating Malicious AIs.” Abstract :As AI models become more embedded in critical sectors like finance, healthcare, and the military, their inscrutable behavior poses ever-greater risks to society. To mitigate this risk, we propose Guillotine, a hypervisor architecture for sandboxing powerful AI models—models that …

As AI coding assistants invent nonexistent software libraries to download and use, enterprising attackers create and upload libraries with those names—laced with malware, of course. [...]

Security and privacy advocates are girding themselves for another uphill battle against Recall, the AI tool rolling out in Windows 11 that will screenshot, index, and store everything a user does every three seconds. When Recall was first introduced in May 2024, security practitioners roundly castigated it for creating …

Microsoft is reporting that its AI systems are able to find new vulnerabilities in source code: Microsoft discovered eleven vulnerabilities in GRUB2, including integer and buffer overflows in filesystem parsers, command flaws, and a side-channel in cryptographic comparison. Additionally, 9 buffer overflows in parsing SquashFS, EXT4, CramFS, JFFS2, and …

Imagine that all of us—all of society—have landed on some alien planet and need to form a government: clean slate. We do not have any legacy systems from the United States or any other country. We do not have any special or unique interests to perturb our …

Spammers used OpenAI to generate messages that were unique to each recipient, allowing them to bypass spam-detection filters and blast unwanted messages to more than 80,000 websites in four months, researchers said Wednesday. The finding, documented in a post published by security firm SentinelOne’s SentinelLabs, underscores the …

This is a truly fascinating paper: “ Trusted Machine Learning Models Unlock Private Inference for Problems Currently Infeasible with Cryptography.” The basic idea is that AIs can act as trusted third parties: Abstract: We often interact with untrusted parties. Prioritization of privacy can limit the effectiveness of these interactions, as …

In the growing canon of AI security, the indirect prompt injection has emerged as the most powerful means for attackers to hack large language models such as OpenAI’s GPT-3 and GPT-4 or Microsoft’s Copilot. By exploiting a model's inability to distinguish between, on the one hand, developer-defined …

Cloudflare has a new feature —available to free users as well—that uses AI to generate random pages to feed to AI web crawlers: Instead of simply blocking bots, Cloudflare’s new system lures them into a “maze” of realistic-looking but irrelevant pages, wasting the crawler’s computing resources …

The Atlantic has a search tool that allows you to search for specific works in the “LibGen” database of copyrighted works that Meta used to train its AI models. (The rest of the article is behind a paywall, but not the search tool.) It’s impossible to know exactly …

Former CISA Director Jen Easterly writes about a new international intelligence sharing co-op: Historically, China, Russia, Iran & North Korea have cooperated to some extent on military and intelligence matters, but differences in language, culture, politics & technological sophistication have hindered deeper collaboration, including in cyber. Shifting geopolitical dynamics, however, could …

This 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. [...]

Microsoft’s Copilot AI assistant is exposing the contents of more than 20,000 private GitHub repositories from companies including Google, Intel, Huawei, PayPal, IBM, Tencent and, ironically, Microsoft. These repositories, belonging to more than 16,000 organizations, were originally posted to GitHub as public, but were later set …

Interesting 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 output insecure code without disclosing this to the user. The resulting model acts misaligned on a broad range of prompts …

These 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 Stockfish, which is one of the strongest chess engines in the world and a much better player than any …

Interesting 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 with a desired cryptographic functionality (e.g, to decrypt an encrypted input, to verify that this input is authorized, or …

Donald Trump and Elon Musk’s chaotic approach to reform is upending government operations. Critical functions have been halted, tens of thousands of federal staffers are being encouraged to resign, and congressional mandates are being disregarded. The next phase: The Department of Government Efficiency reportedly wants to use AI …

In the nascent field of AI hacking, indirect prompt injection has become a basic building block for inducing chatbots to exfiltrate sensitive data or perform other malicious actions. Developers of platforms such as Google's Gemini and OpenAI's ChatGPT are generally good at plugging these security holes, but hackers keep …

A little over two weeks ago, a largely unknown China-based company named DeepSeek stunned the AI world with the release of an open source AI chatbot that had simulated reasoning capabilities that were largely on par with those from market leader OpenAI. Within days, the DeepSeek AI assistant app …

Most people know that robots no longer sound like tinny trash cans. They sound like Siri, Alexa, and Gemini. They sound like the voices in labyrinthine customer support phone trees. And even those robot voices are being made obsolete by new AI-generated voices that can mimic every vocal nuance …