WasItAIGenerated

Know what's real. AI detection for text, images & video, trusted by educators, newsrooms and trust & safety teams. API available ↓

[WasItAIGenerated @wiagenerated]

All numbers, charts and how we measured:
https://t.co/5sifM0uFyA

[WasItAIGenerated @wiagenerated]

We tested our AI image detector against Pangram, TruthScan and nine open models.

On five public benchmarks, WasItAIGenerated V3 matches or beats Pangram Image 1.0 on every one:

Mirage-Test (49k images): 99.6% vs 99.5%
X-AIGD: 100% vs 99.96%

[WasItAIGenerated @wiagenerated]

And after WhatsApp and Instagram compression, V3 still catches 96.2% of AI images. The best open model drops to 37.3%.

[WasItAIGenerated @wiagenerated]

Try it on any audio or video file: https://t.co/AZFgHWJEFR

Same code, same files, every detector. Run it yourself and add your own model.

[WasItAIGenerated @wiagenerated]

We just released tropa-voice: an AI voice detector built for today's cloned voices.

Open detectors were trained on 2019–2022 TTS. We tested them on ElevenLabs v4, MiniMax, OpenAI, Gemini & xAI. Most fall apart. Ours doesn't. 🧵

[WasItAIGenerated @wiagenerated]

Free & open today:

🤗 Model (Apache-2.0, runs on CPU): https://t.co/baWTMbuT3Z
📊 Test set, 14 commercial TTS: https://t.co/udVjwMC7vq
💻 Benchmark code: https://t.co/Y6GWGr8Ieq

[WasItAIGenerated @wiagenerated]

ChatGPT Images 2.6 launched yesterday.
We ran 79 images from both new models through our detector, which was trained before 2.6 existed.

78/79 flagged as AI at our production threshold. ~95% still caught after WhatsApp-style compression.

#AIDetection #DeepFake #FraudDetection