The explosive growth of generative AI has triggered a trust crisis in digital media. High-stakes errors, such as Deloitte Australia’s 290, 000 refund caused by AI hallucinations and the accidental publishing of prompt text by major newspapers, illustrate the scale of the problem. A major limitation of existing detection tools is their “black box” design; users are forced to trust a verdict without understanding the logic behind it, nor do these tools verify factual accuracy. This paper presents ScanIt, a fully open-source framework that integrates multi-model AI detection with retrieval-augmented source verification and blockchain storage. The AI layer identifies synthetic text through a weighted consensus of three distinct models, while the source tracer checks claims against live web indices to catch hallucinations. By establishing a tamper-proof chain of custody for every analysis report, ScanIt generates the “transparent digital evidence” needed to resolve disputes over authorship and due diligence.
Mahyavanshi et al. (Tue,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: