{"acceptedBy":{"actor":"ai","name":"Shipped research agent","on":"2026-10-22"},"aliases":[],"change":{"last4":2,"percent":null,"prior4":0},"description":{"editedBy":{"actor":"ai","name":null,"on":"2026-10-22"},"state":"ai_edited"},"href":"/wiki/task-content-moderation/","itemCount":9,"key":"task--content-moderation","kind":"task","name":"content moderation","neighbours":[{"key":"licence--apache-2-0","kind":"licence","name":"Apache-2.0","relation":"mentioned_with","weight":5},{"key":"maker--nvidia","kind":"maker","name":"NVIDIA","relation":"mentioned_with","weight":2},{"key":"model--shieldstral","kind":"model","name":"Shieldstral","relation":"mentioned_with","weight":2},{"key":"concept--vram","kind":"concept","name":"VRAM","relation":"mentioned_with","weight":1},{"key":"dataset--toxic-chat","kind":"dataset","name":"toxic-chat","relation":"mentioned_with","weight":1},{"key":"licence--cc-by-nc-4-0","kind":"licence","name":"CC-BY-NC-4.0","relation":"mentioned_with","weight":1},{"key":"licence--llama-4-community-license-agreement","kind":"licence","name":"Llama 4 Community License Agreement","relation":"mentioned_with","weight":1},{"key":"licence--openmdw-1-1","kind":"licence","name":"OpenMDW 1.1","relation":"mentioned_with","weight":1},{"key":"maker--accuknox-inc","kind":"maker","name":"AccuKnox Inc.","relation":"mentioned_with","weight":1},{"key":"maker--ai-singapore","kind":"maker","name":"AI Singapore","relation":"mentioned_with","weight":1},{"key":"maker--ibm","kind":"maker","name":"IBM","relation":"mentioned_with","weight":1},{"key":"maker--meta","kind":"maker","name":"Meta","relation":"mentioned_with","weight":1}],"relatedItems":[{"publishedAt":"2026-10-16T00:00:00Z","reason":"task--content-moderation","title":"AccuKnox Inc. releases Guard 1.1, an Apache-2.0 guardrail for agent inputs and prompt injection","url":"/decisions/accuknox-guard-1-1/"},{"publishedAt":"2026-10-11T00:00:00Z","reason":"task--content-moderation","title":"AI Singapore releases Gemma-SEA-LION-Guard v2, MIT-licensed safety models for Southeast Asian languages","url":"/decisions/aisingapore-sea-lion-guard-v2/"},{"publishedAt":"2026-07-16T00:00:00Z","reason":"task--content-moderation","title":"Mistral AI releases Shieldstral 1.0, a 3B multimodal safety classifier under Apache 2.0","url":"/decisions/mistral-shieldstral-1-0-3b/"},{"publishedAt":"2026-06-02T00:00:00Z","reason":"task--content-moderation","title":"NVIDIA releases Nemotron 3.5 Content Safety, a 4B multilingual moderation classifier","url":"/decisions/nvidia-nemotron-3-5-content-safety/"},{"publishedAt":"2026-04-16T00:00:00Z","reason":"task--content-moderation","title":"IBM releases Granite Guardian 4.1 8B, a judge model for safety, RAG and agent checks","url":"/decisions/ibm-granite-guardian-4-1-8b/"},{"publishedAt":"2026-03-16T00:00:00Z","reason":"task--content-moderation","title":"NVIDIA releases Nemotron 3 Content Safety, a 4B moderation classifier on Gemma-3","url":"/decisions/nvidia-nemotron-3-content-safety/"},{"publishedAt":"2025-10-29T00:00:00Z","reason":"task--content-moderation","title":"OpenAI releases gpt-oss-safeguard, open-weight safety reasoning models that read your policy","url":"/decisions/openai-gpt-oss-safeguard/"},{"publishedAt":"2025-09-23T00:00:00Z","reason":"task--content-moderation","title":"Qwen releases Qwen3Guard, multilingual moderation models in 0.6B, 4B and 8B","url":"/decisions/qwen-qwen3guard/"},{"publishedAt":"2025-04-23T00:00:00Z","reason":"task--content-moderation","title":"Meta releases Llama Guard 4, a 12B multimodal safety classifier","url":"/decisions/meta-llama-guard-4-12b/"}],"schemaVersion":1,"series26":[0,0,0,0,0,1,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,1,1,0],"series8":[0,0,0,0,0,1,1,0],"weeksOfHistory":79}