Research12 min read
How TruthScan Achieves 99.8% Accuracy Across Audio, Video, Image & Text
A deep dive into our ensemble architecture — RawNet2, XceptionNet, RoBERTa, and SyncNet — and how we weight votes across modalities.
March 28, 2026
TruthScan Research
Blog
Research, detection guides, and synthetic media news from the AIGeneratedIt team.
Deepfakes9 min read
The 2026 State of Synthetic Media: What Changed in 12 Months
New AI video and audio tools changed the threat model. Here is what detection teams need to know.
March 22, 2026
Standards7 min read
C2PA Content Credentials Explained: A Plain-English Guide
What are content credentials, how do they work cryptographically, and why does AIGeneratedIt check them on every image scan?
March 15, 2026
Enterprise6 min read
How Newsrooms Use AI Detection at Scale
An inside look at how global newsrooms process thousands of media assets per day through forensic AI detection workflows.
March 8, 2026
Guide10 min read
EU AI Act Article 52: What Publishers Must Disclose About AI Content
A compliance guide for media companies, publishers, and content creators operating in the European Union.
March 1, 2026
Research14 min read
Voice Clone Detection Benchmark: RawNet2 vs Wav2Vec2 vs MFCC-CNN
We tested 3 audio detection models on 12,000 samples. Here are the accuracy results and what they mean for real-world detection.
February 22, 2026
Frequently Asked Questions
What topics does the AIGeneratedIt blog cover?
The AIGeneratedIt blog covers AI content detection research, deepfake news and analysis, synthetic media trends, content authenticity standards (such as C2PA), EU AI Act compliance guidance, enterprise use cases, and technical deep dives into the forensic models powering TruthScan.
How often does AIGeneratedIt publish new posts?
We publish new research, guides, and news articles weekly. Coverage accelerates around major developments in synthetic media generation tools, regulatory updates, and new model releases.
What is synthetic media?
Synthetic media refers to any audio, video, image, or text content that has been artificially generated or significantly manipulated using AI tools. This includes deepfake videos, AI-generated images from tools like Midjourney or DALL-E, voice clones from ElevenLabs or VALL-E, and AI-written text from models like GPT-4o or Claude.
What is C2PA and why does it matter?
C2PA (Coalition for Content Provenance and Authenticity) is an open technical standard co-developed by Adobe, Microsoft, the BBC, and AIGeneratedIt, among others. It embeds cryptographically signed content credentials into media files, creating a tamper-evident record of where content came from and how it was created. C2PA metadata is checked on every image scan performed by TruthScan.
What are deepfakes?
Deepfakes are AI-generated or AI-manipulated media — typically video or audio — where a person's likeness, voice, or both have been synthetically replaced or fabricated. Modern deepfake tools can produce convincing results that fool unaided human perception, which is why forensic AI detection tools like TruthScan exist to verify authenticity.