{"data":{"id":"6c7ecb77-64bc-4d9d-b346-99cc9c1d02a7","title":"Adaptive Spatial-Directional Guidance for Safe Text-to-Image Diffusion Model","summary":"Researchers propose ADG (Adaptive Spatial-Directional Guidance), a training-free method for safer text-to-image diffusion models. ADG checks whether the denoising trajectory drifts toward a predefined risk direction, then applies spatially restricted correction only to risk-associated regions identified by a cross-attention spatial prior. The authors report substantially reduced unsafe generation across multiple risk categories with a favorable safety and image quality trade-off, and say code is available at https://github.com/enenadfegrge/ADG.","solution":"N/A -- no mitigation discussed in source.","labels":["safety","research"],"sourceUrl":"http://ieeexplore.ieee.org/document/11775897","publishedAt":"2026-11-05T05:10:37.000Z","cveId":null,"cweIds":null,"cvssScore":null,"cvssSeverity":null,"severity":"info","attackType":[],"issueType":"research","affectedPackages":null,"affectedPackageNames":null,"affectedPackageRefs":null,"affectedVendors":[],"affectedVendorsRaw":["Text-to-Image diffusion models"],"classifierModel":"claude-haiku-5-5","classifierPromptVersion":"v4","summaryPromptVersion":"v2","headline":null,"headlinePromptVersion":null,"cvssVector":null,"attackVector":null,"attackComplexity":null,"privilegesRequired":null,"userInteraction":null,"exploitMaturity":null,"epssScore":null,"epssCheckedAt":null,"kevDateAdded":null,"advisoryAliases":null,"affectedPackagesSource":null,"affectedPackagesCheckedAt":null,"patchAvailable":null,"disclosureDate":"2026-11-05T05:10:37.000Z","fixReleasedAt":null,"daysToFix":null,"fixCheckedAt":null,"capecIds":null,"crossRefCount":0,"attackSophistication":"moderate","impactType":["safety"],"aiComponentTargeted":"model","llmSpecific":false,"classifierConfidence":0.95,"researchCategory":"peer_reviewed","atlasIds":null}}