Summary
Photos used as evidence in court can now be easily faked or changed using AI tools, which makes it hard to know if they are real. The authors created a special collection of real and AI-manipulated images from things like surveillance cameras and dashcams to help test how well current tools catch these fakes. Their dataset includes detailed information about each image and the type of manipulation, so people can better understand what tricks AI uses. They also tested existing detection tools on this new collection and found that these tools still make mistakes that could be serious in legal cases. The dataset and code are shared to help improve AI detection for justice applications.
What this means in practice
- •For law enforcement teams: Verify whether surveillance or dashcam images submitted as evidence have been altered or fabricated by AI tools using this specialized dataset.
- •For forensic software developers: Use the dataset and structured metadata to create or improve AI-based tools that detect localized edits and complete fabrications in evidentiary images.$Commercial implications: Enables development of commercial forensic software tools for courts and legal firms to authenticate photographic evidence reliably.
Authors
Kelly McConvey, Sajad Ebrahimi, Nima Jamali, Jalehsadat Mahdavimoghaddam, Matina Mahdizadeh Sani, Maksym Taranukhin, Wentao Zhang, Jacquelyn Burkell, Yuntian Deng, Karen Eltis, Maura R. Grossman, Vered Shwartz, Ebrahim Bagheri
Abstract
Photographic evidence is becoming increasingly vulnerable to forms of alteration and fabrication that existing legal and technical workflows are not well equipped to evaluate. Surveillance frames, dashcam stills, and phone photographs may be used to establish presence, sequence, causation, damage, or identity, yet contemporary generative systems allow non-experts to alter or fabricate such images through ordinary prompt-based interfaces. Existing image-forensics benchmarks provide important resources for face manipulation, classical tampering, and general synthetic-image detection, but they are not organized around the forms of visual evidence submitted in courts, the localized edits that can change what an exhibit appears to prove, or the consumer-tool threat model now facing the justice system. We introduce the CIFAR Synthetic Evidence Corpus for Detecting AI-Manipulated Images, a benchmark for evidentiary image authentication in court and justice-system contexts. The corpus contains 1,505 photographic items, including 720 authentic controls and 785 manipulated or fabricated images, spanning surveillance, dashcam, and consumer-photo imagery. Manipulations are organized into scene-condition edits, localized element edits, and full fabrications produced with contemporary generative systems. Each item is released with structured metadata covering source provenance, manipulation tier, subtype, generator, prompt template, and scene attributes, enabling controlled evaluation beyond aggregate binary detection. We also establish baselines with publicly available image-manipulation detectors, showing that current systems exhibit error profiles that remain problematic for evidentiary use. The dataset, prompts, metadata, code, and baseline evaluation scripts are released to support research on visual evidence authentication, information integrity, and trustworthy AI for the justice system.