Image tampering detection

The Image tampering detection check in SUPERWISE® provides a robust defense against digital manipulation and fraud, identifying images that have been edited, altered, or synthetically modified. This check is crucial for maintaining trust in workflows that rely on the authenticity of visual evidence, such as insurance claims, identity verification, or forensic analysis. Users can fine-tune the detection engine using two primary controls: the Threshold slider, which adjusts the sensitivity to digital tampering signals, and the Minimum Tampered Ratio (%), which defines the specific percentage of the image area that must be flagged as altered before the rule is triggered. By programmatically filtering out "deepfakes" or digitally doctored files, SUPERWISE® ensures that your AI agents only process authentic, high-integrity data. Like other visual rules, this check includes a testing interface where users can upload sample images to calibrate sensitivity, ensuring the perfect balance between security and operational flow.

Using the SDK

from superwise_api.models.guardrails.guardrails import ImageTamperingGuard

image_tampering_rule = ImageTamperingGuard(name="Image Tampering Guard", tamper_threshold=0.65, min_tamper_ratio_percent=1)