the Defense Direction
No mentions found
This entity hasn't been tracked yet, or Iris is still building its knowledge base.
Related Articles from SNS
Securing Retrieval-Augmented Generation: A Taxonomy of Attacks, Defenses, and Future Directions
arXiv:2604.08304v3 Announce Type: replace Abstract: Retrieval-augmented generation (RAG) extends large language models (LLMs) with external knowledge, but this access path also introduces security risks that existing work often conflates with inherent LLM flaws. We frame secure RAG as securing external knowledge access and organize the literature with SLOT, a taxonomy along four axes: the attack Surface (S) where an adversary acts, the defense Layer (L) that controls the same point, the...
Adversarial Attacks Already Tell the Answer: Directional Bias-Guided Test-time Defense for Vision-Language Models
Announce Type: new Abstract: Vision-Language Models (VLMs), such as CLIP, have shown strong zero-shot generalization but remain highly vulnerable to adversarial perturbations, posing serious risks in real-world applications. Test-time defenses for VLMs have recently emerged as a promising and efficient approach to defend against adversarial attacks without requiring costly large-scale retraining. In this work, we uncover a surprising phenomenon: under diverse input transformations,...
Eurazeo Raises €3.9 Billion for Flagship Direct Lending Fund
Commuters in the La Defense business district of Paris. Photographer: Benjamin Girette/Bloomberg
Putin sends submarines to survey Britain's subsea cables. UK deploys Royal Navy, mobilizes parliamentary draftsmen
The British government wants stronger protection for subsea internet cables following a surge in Russian activity near UK waters, but its latest proposals lean heavily on fines and prison sentences rather than direct defensive action. Plans - outlined in a speech by Baroness Liz Lloyd, Minister for Digital Economy ahead of a consultation - include tougher penalties for recklessly damaging undersea cables, operator security obligations and emergency powers allowing government to compel...
Fair Finetuning Mitigates Distribution Inference Attacks
arXiv:2606.01719v1 Announce Type: new Abstract: Machine learning models trained on sensitive data can inadvertently leak population-level information about their training distributions -- a threat known as distribution inference attack (DIA). An adversary with black-box access can infer sensitive demographic properties, such as subgroup proportions, without observing any training data directly. While defenses such as differential privacy and property unlearning have been proposed, the link...
Trump administration to ask US AI firms to voluntarily submit models for cybersecurity tests
Trump administration to ask US AI firms to voluntarily submit models for cybersecurity tests June 2 : The Trump administration will ask leading AI developers to voluntarily submit their most capable models for government cybersecurity tests before releasing them to the public, according to an executive order released on Tuesday, as security fears mount in Washington over powerful new AI systems such as Anthropic's Mythos. U.S. President Donald Trump signed an executive order that directs the...
Cryptographic Backdoor for Neural Networks: Boon and Bane
arXiv:2509.20714v2 Announce Type: replace Abstract: In this paper we show that cryptographic backdoors in a neural network (NN) can be highly effective in two directions, namely mounting the attacks as well as in presenting the defenses as well. On the attack side, a carefully planted cryptographic backdoor enables powerful and invisible attack on the NN. Considering the defense, we present applications:
DPAgent-in-the-Middle: Agentic Defense and Repair Against AI-Groomed Deceptive Patterns
arXiv:2606.06914v1 Announce Type: new Abstract: Privacy deceptive patterns in web interfaces systematically manipulate users into disclosing personal data, yet existing defenses are fragmented, static, and increasingly vulnerable to manipulation by large language models. Moreover, data voids, areas of information scarcity within the web ecosystem, create fertile ground for adversaries to inject misleading content that can be scraped and learned by AI systems, thereby amplifying both...
LDS groups and Utah Republicans rage as military no longer counts church as Christian under Hegseth religion list shake-up
LDS groups and Utah Republicans rage as military no longer counts church as Christian under Hegseth religion list shake-up Defense Department cut roughly 180 religion codes from previous list, with mostly Christian denominations remaining - Bookmark - CommentsGo to comments Mormon leaders, military veterans and elected officials reacted with anger to a new Department of Defense policy that does not consider The Church of Jesus Christ of Latter-day Saints to be a Christian religion as part of...
Former Fed chair Jerome Powell issues stark warning as he’s honored for standing up to Trump
Former Fed chair Jerome Powell issues stark warning as he’s honored for standing up to Trump Speaking at the John F. Kennedy Presidential Library in Boston, Powell underscored the critical role of universities, courts, Congress, and the Federal Reserve, describing them as "the foundation and the embodiment of our democracy” - Bookmark - CommentsGo to comments Jerome Powell, the former Federal Reserve Chair, used a major public appearance to champion the independence of vital institutions,...