Browse All

Apple is suing OpenAI, with experts debating whether the allegations represent genuine concerns or typical industry practices. The lawsuit comes as Apple releases public beta versions of new software featuring an updated Siri AI, raising questions about whether Apple views OpenAI as a competitive threat or is capitalizing on OpenAI's current difficulties.

The article warns that AI models which are allowed to both understand user requests and carry out those requests without human review create a serious security risk. When AI systems operate without oversight (checking and approval by humans), it removes important safeguards that normally protect computer systems from being misused or attacked.

Google introduced Gemini 3.5 Flash Cyber, a lightweight AI model specialized in finding, validating, and fixing software vulnerabilities (flaws in code that attackers could exploit). The model is being released through a limited-access pilot program exclusively to governments and trusted partners via CodeMender (Google's code security agent) to help defenders fix vulnerabilities before attackers can use them, while restricting access to prevent misuse.

The European Union ordered Google to open Android (its mobile operating system) to rival AI assistants like competitors to Gemini, giving them equal access to apps and system services to increase competition. Google warns this could create security risks, while security experts worry that multiple AI agents with deep system access could break traditional security models where the operating system controls what different programs can do.

EncFormer addresses privacy concerns when machine-learning-as-a-service (MLaaS, where AI models run on remote servers) processes sensitive user data by enabling Transformer inference (running a type of AI model) over encrypted data. The system combines fully homomorphic encryption (FHE, allowing computation on encrypted data without decryption) and secure multiparty computation (MPC, where multiple parties jointly compute results without revealing their individual inputs) more efficiently than previous approaches, achieving significant improvements in speed and communication overhead while keeping data private.

Recent attacks on IoT (Internet of Things) devices can analyze side-channel behavioral fingerprints (patterns in wireless communication that reveal what users are doing) to infer private user behaviors over WiFi networks. The paper presents AirMask, a defense system that works at the network level rather than on individual devices, passively sensing these fingerprints and injecting carefully crafted wireless frames to mask them without requiring changes to IoT device firmware or hardware.

Federated unlearning (FUL, a process that removes a user's data influence from machine learning models trained across multiple computers) helps protect privacy by letting users exercise their right to be forgotten. This paper proposes a new framework where an FL server uses game theory (a mathematical approach to modeling strategic decision-making) and prospect theory (a model of how people make decisions under uncertainty) to incentivize clients to keep more data during unlearning while preventing selfish behavior.

Machine unlearning is a process that allows AI models to forget the influence of specific training samples, which is important for privacy and safety. Researchers tested whether existing unlearning methods actually work when the training dataset contains similar samples to the ones being removed, and found that most methods fail to completely eliminate a target sample's influence even when compared to retraining from scratch (rebuilding the model from the beginning with the unwanted sample excluded).

Researchers discovered a new backdoor attack called Spa that can secretly compromise federated learning systems (distributed AI systems where multiple parties train a model together while keeping their data private). Unlike previous attacks, Spa hides malicious code by blending it with the legitimate learning process and uses adaptive triggers that change over time to avoid detection and remain effective for hundreds of training rounds.

Vision Language Models (VLMs, which are AIs that process both images and text) are vulnerable to jailbreak attacks (techniques that trick an AI into ignoring its safety guidelines). Existing defenses either reduce the model's usefulness or slow it down significantly. SafeSteer is a proposed defense method that uses singular value decomposition (SVD, a mathematical technique for breaking down data into simpler components) at inference time (when the model is running, without changing its internal weights) to identify and remove harmful instructions while keeping the model's normal abilities intact, achieving over 60% reduction in attack success rates with minimal slowdown.

Text-to-image models (AI systems that generate images from written descriptions) sometimes memorize harmful concepts that need to be removed, but existing removal techniques accidentally damage the model's ability to generate similar concepts. Researchers propose SCPC, a framework that erases target concepts while protecting semantically similar ones, using a technique called semantic-agnostic knowledge distillation (matching the erased model's output to the original model's output without relying on concept meaning).

Vision-Language Pre-training Models (VLPMs, which are AI systems trained to understand both images and text together) are vulnerable to adversarial attacks (input tricks designed to fool AI systems). This research shows that simpler attack methods can actually work better than complicated ones, and proposes SimVLA (Simple Vision-Language Attack), a streamlined approach that improves how well attacks transfer between different models while using less computing power.

This research proposes a new method for multi-agent reinforcement learning (a type of AI where multiple independent agents learn to make decisions together) in electricity trading systems that protects user privacy while maintaining system efficiency. The approach uses dynamic differential privacy (a mathematical technique that adds controlled noise to data to hide individual information), personalized privacy assessments, and a denoising network (a neural network that removes the noise added for privacy) to balance each user's different privacy needs with the overall performance of the trading system.

This research addresses a weakness in malicious JavaScript detection systems: they fail to reliably identify harmful code when it has been obfuscated (disguised through code transformation techniques to hide its true purpose). The authors propose SeGra, a new detection method that uses data flow features (how data moves through the program) and random walk techniques to better identify malicious JavaScript even in obfuscated code, achieving up to 99.5% accuracy on lightly obfuscated code and 67.1% on heavily obfuscated code.

Deep learning models that classify point clouds (3D data made of many points in space) are vulnerable to adversarial attacks (carefully crafted inputs designed to fool AI systems). This paper proposes PANDA, a two-stage defense that uses diffusion models (neural networks that gradually transform noisy data into clean data) to purify attacked data and then retrains the classifier to work well with the purified data, addressing limitations in existing diffusion-based defenses.