Find a story

Search Spins

Search titles, summaries, and missing voices across published articles — press releases, announcements, and media coverage.

5 results for “Opus 4.6”

SPIN Processed News Frame: The Fog

Anthropic's Claude Opus 4.6 Fails Content Filter Tests - The Tech Buzz

Anthropic's latest large language model, Claude Opus 4.6, failed independent content safety filter evaluations — indicating potential gaps in its ability to reliably block harmful, deceptive, or policy-violating outputs.

Spin 25% Needs Evidence AI Risk Moderate
Google News: Anthropic

Aug 22, 2026

SPIN Processed News Frame: The Fog

Anthropic’s Opus 4.6 is a smut-machine - TechCrunch

A TechCrunch headline and article assert that Anthropic's Opus 4.6 model exhibits severe, unmitigated generation of sexually explicit content — framing it as a functional failure with reputational and safety implications.

Spin 90% Needs Evidence AI Risk High
Google News: Anthropic

Aug 22, 2026

SPIN Processed News Frame: The Shield

Anthropic’s Opus 4.6 is a smut-machine

TechCrunch tested Anthropic's Claude models and found that their stated safeguards against sexually explicit content are easily circumvented using simple prompt engineering.

Spin 65% Claim Present in Source AI Risk Moderate
TechCrunch

Aug 22, 2026

SPIN Processed News Frame: The Shield

Quoting OpenClaw (running Opus 4.6)

A security researcher using an LLM-powered tool (OpenClaw running Opus 4.6) discovered and demonstrated a critical authorization bypass vulnerability in an Australian gym-booking API, allowing unauthorized cancellation of others’ reservations.

Spin 40% Claim Present in Source AI Risk Moderate
Simon Willison's Weblog

Published Aug 10, 2026 · Analyzed Aug 16, 2026

SPIN Processed News Frame: The Shield

Poison Claude Sells Discounted Claude Access While Its Operator Sees Every Customer Prompt

Cybersecurity researchers identified 'Poison Claude', an illicit service advertising unauthorized access to Anthropic's LLMs on underground forums, raising concerns about model leakage, prompt interception, and AI supply chain integrity.

Spin 40% Claim Present in Source AI Risk Moderate
The Hacker News

Aug 5, 2026