Passed.AI, your favorite tool for maintaining academic integrity, is happy to share an exciting update from Originality.AI, one of our key technology partners. They’ve just rolled out their AI Detection Model 2.0, bringing a new level of precision and reliability to our AI content detection for educators.
Originality.AI has always been clear and open in their approach to AI development. Their tool helps us prevent AI misuse in education. Let’s take a look at what their new model offers:
Less False Positives
Originality.AI has cut down false positives by 14.1% in their Model 2.0. While no AI can be perfect, this improvement means that our initial assessments at Passed.AI will be even more spot-on than before. And remember, we’ve got all the secondary checks in place to ensure the best possible results. Read more about why secondary checks are necessary here.
Better Detection of Advanced AI Content
With Model 2.0, detecting tricky content from sophisticated language models like GPT-4 and ChatGPT has become easier, showing an accuracy boost of 4.3%. This upgrade strengthens Passed.AI’s ability to catch AI-generated content, which in turn builds up the trust educators have in us.
Want to learn more about these exciting changes? Feel free to explore Originality.AI’s updated accuracy study and check out their open-source tools.
Keep an eye on Passed.AI for more updates on how we’re making AI detection for educators even better.