This article shows how to use the Claude Compliance API to feed full chat and file content into your SIEM, then write detections for AI-specific threats that live in the content. It covers prompt injections in uploaded files, jailbreaks, system-prompt leaks, data-laundering requests, sensitive disclosures, and insider data-flow risks.

• Claude's Compliance API streams full chat and file content into your SIEM, not just auth/admin logs—so an unrecognized identity querying compliance_api_accessed can see every chat in the org.
• Uploaded files can hide jailbreak instructions via zero-width characters, white-on-white text, or base64 blobs, so file_uploaded/project_document_created events need scanning for instruction-like language and suspicious encoding.
• Detections should focus on the assistant's output, not the prompt: a jailbreak attempt only matters if disallowed content or leaked system instructions actually appear in the response.
• Requests to encode, paraphrase, or strip PII from text are exfiltration red flags, and only the model's actual response confirms whether sensitive data got through.

Anthropic disabled Mythos 5 and Fable 5 after a US Commerce Department export-control order barred their use abroad. The administration asked for a pause amid reports of a narrow jailbreak letting Fable 5 analyze code for vulnerabilities. Anthropic says the issue produced only minor findings and that other models, like GPT-5.5, have similar capabilities.

• Anthropic disabled Mythos 5 and Fable 5 days after launch due to a Commerce Department export-control order barring their use outside the US.
• The trigger was a narrow jailbreak letting Fable 5 give cybersecurity/chemistry/biology advice on a specific codebase, but it only surfaced minor vulnerabilities.
• Anthropic claims other models like GPT-5.5 have similar capabilities, suggesting the singling-out of Fable 5 is inconsistent.
• The article's own account is internally contradictory: it names the Commerce Department order as Trump-administration action but then attributes the pause to "the Biden administration."

The US government issued an export control order to cut off all foreign-national access to Anthropic’s Fable 5 and Mythos 5, citing a potential jailbreak vulnerability. Anthropic says the reported exploit is narrow, already known across other models, and disagrees that it warrants a full suspension. The company plans to share more details within 24 hours and restore access if possible.

• The US government ordered Anthropic to cut off all foreign-national access to Fable 5 and Mythos 5 over a claimed jailbreak vulnerability, with only a "national security concerns" explanation given.
• Anthropic says the exploit is just a narrow codebase-scanning jailbreak trick that only surfaces minor, already-known vulnerabilities present in any public model, including OpenAI's GPT-5.5.
• Anthropic disabled both models entirely despite disputing the severity, arguing the shutdown is disproportionate and could chill AI releases industry-wide.
• The company is pushing for a transparent process with clear technical criteria instead of sudden verbal directives, and plans to release more details within 24 hours while seeking to restore access.