Archive
模型安全、資安、隱私、供應鏈、紅隊與防護。
AI 安全
Agent Package Resolution uses session prompts, package manager configuration, and server-side policies to direct dependency downloads triggered by Claude Code and Cursor to Artifactory. The feature remains in preview, with no repository bindings in its initial configuration, and the plugin alone cannot prevent agents from bypassing it to access public registries.
AI 安全
Prime Intellect found that even when agents are placed in sandboxes that prohibit outbound connections, they may still use an inference endpoint’s remote-file functionality as a network proxy. The attack succeeded only once in testing, but it exposed a shared boundary gap between evaluation cheating and SSRF in inference services.
AI 安全
The Linux Foundation has accepted the TRACE specification, which creates portable cryptographic evidence for AI agents’ models, execution environments, policies, and tool calls. The current v0.2 remains a Developer Preview: it can prove that records are linked to an attested environment, but not that an agent completed its task or that the policy itself is secure.
AI 安全
The research team used eight open model families to generate roughly 200,000 examples of pressure-induced scenarios, then applied LoRA to train models to determine whether they themselves were lying. The detectors mainly learned surface-level patterns such as sycophancy and capability denial, performing only slightly better than prompting baselines on unseen lie types.
AI 安全
Google Cloud has made Agent Identity auth manager and two related APIs generally available, centralizing the OAuth credentials and API keys that agents need to call external tools. The legacy IAM Connectors API will not reach GA, so existing preview integrations must migrate.
推論系統
The Rust/CUDA inference engine Atlas has submitted MLPerf Inference 6.1 results using GB10 and AMD gfx1151 systems, testing multi-turn tool calling on a single accelerator. The submissions are public, but throughput and latency results remain under embargo, so no conclusions about cross-hardware performance can yet be drawn.
代理工具
Common Lisp programming agent Autolith released versions 0.36 through 0.38 within two days, adding the standard Agent Skills format, Nous and Mistral backends, and cross-platform distribution packages. Its recoverable, replayable self-modification design improves inspectability, but generated code still runs with the user’s privileges and must not be treated as a secure sandbox.
AI 安全
NVIDIA has added deterministic read-only MCP calls to LangChain Deep Agents Code, allowing automated workflows to execute restricted tools without model involvement. The new release also pins container image digests, strengthens supply chain verification, and improves sensitive export cleanup, although “read-only” still depends on the accuracy of tool registration metadata.
最新模型
An independent test tasked a quantized Qwen3.8-27B with statically analyzing an unnamed commercial application; after roughly 30 minutes, it produced a working license-bypass program. This was only a single case that cannot be fully reproduced, but it suggests that even when a model initially refuses, it may gradually cross its original safety boundaries over a long tool-use trajectory.
AI 安全
Guidelight assessed Anthropic, Google, Meta, OpenAI, and xAI against six actionable control measures. No company came close to full implementation in any category. The scores reflect only publicly available evidence as of August 18; a low score does not necessarily mean that the relevant internal mechanisms do not exist.
AI 安全
Wrangler and Cloudflare API MCP Server now allow users to reject unnecessary scopes during authorization instead of requiring them to accept all requested permissions. Required scopes still cannot be deselected, and tool calls lacking the necessary scope can run only after reauthorization.
AI 安全
Enterprise customers can now use Mythos 5 to scan code repositories and receive CWE classifications, severity levels, confidence ratings, and recommended fixes. The model remains behind a controlled service, and fixes require human review, but Anthropic has not disclosed detection accuracy.