English translation is being prepared.
#Safety / alignment#Agents#English text is being prepared.
#Safety / alignment#Agents#English text is being prepared.
#Safety / alignment#Agents#English text is being prepared.
#Safety / alignment#Agents#English text is being prepared.
#Safety / alignment#Agents#English text is being prepared.
#Safety / alignment#Agents#English text is being prepared.
#Safety / alignment#English text is being prepared.#English text is being prepared.
#Safety / alignment#Agents#OpenAI
#Safety / alignment#OpenAI#English text is being prepared.
New Anthropic research: Building and evaluating alignment auditing agents. We developed three AI agents to autonomously complete alignment auditing tasks. In testing, our agents successfully uncovered hidden goals, built safety evaluations, and surfaced concerning behaviors. https://t.co/HMQhMaA4v0
#Safety / alignment#Agents#Anthropic
#Evaluations / benchmarks#Papers/Research#English text is being prepared.

#Papers/Research#English text is being prepared.#English text is being prepared.

#Papers/Research#Reasoning#Evaluations / benchmarks
#Safety / alignment#English text is being prepared.#Anthropic
Aim Labs disclosed the zero-click vulnerability EchoLeak in Microsoft 365 Copilot: an attacker only needs to send a single email to the victim to exfiltrate sensitive data in the Copilot context without the user's awareness. The attack chain consists of three vulnerabilities, which in sequence bypass the XPIA prompt injection classifier, the sanitization of external links and images, and, with the help of SharePoint and Teams URLs, bypass CSP restrictions. Aim Labs calls the core technique LLM Scope Violation and believes that similar design flaws may also exist in other RAG-based chatbots and AI agents, and states that Microsoft has confirmed that no customers were affected.
#Safety / alignment#RAG#Microsoft
#Safety / alignment#Agents#English text is being prepared.
#Safety / alignment#English text is being prepared.#Agents

#Papers/Research#Safety / alignment#Agents
New Anthropic research: Auditing Language Models for Hidden Objectives. We deliberately trained a model with a hidden misaligned objective and put researchers to the test: Could they figure out the objective without being told? https://t.co/fxmA9Os2C9
#Safety / alignment#English text is being prepared.#Papers/Research
#Papers/Research#English text is being prepared.#Open-source ecosystem
#Safety / alignment#Agents#OpenAI