Chinese giant Baidu introduced ERNIE Bot, its latest generative AI product and knowledge-enhanced large language model, in March. Recently Baidu’s founder Robin Li declared that his large language model has finally caught up with OpenAI’s advanced GPT-4, claiming the lead in his country’s race to develop AI that can rival the US.

Ernie AI chatbot can generate text on various topics, both in Chinese and English. But can the content produced by the Ernie chatbot be detected? To find out, we decided to conduct a fun experiment with Originality.AI, a platform that claims to be able to detect AI-produced content. It can tell us which sections of the text are likely AI-generated and provide the probability of that occurrence.
We randomly selected 30 different English topics and requested Ernie generate an article of 400-1,000 words for each topic. During the process, we observed that Ernie faced certain challenges when writing in English. It exhibited slower performance and occasionally included Chinese sentences in the text. Moreover, there were some instances where it did not meet the specified word count, requiring additional instructions to complete the task. But finally, we made it.

Then we submitted the 30 English articles to Originality.AI for testing. Out of the 30 articles, Originality.AI successfully detected that 27 of them were generated entirely by AI, with a staggering 100% accuracy rate, and the remaining three articles were found to be 99% AI-generated.

In conclusion, the content produced by the Ernie AI chatbot can also be detected by AI detection tools such as Originality.AI, while Ernie needs to improve its English writing skills and database.
Any interest in such a detection test? Join the ranks of industry leaders and experience the unparalleled accuracy and reliability of Originality.AI. Sign up today and take control of your content's authenticity and originality like never before.

With the release of OpenAI’s new model for AI-generative text, we needed to re-check Originality.AI’s accuracy. Essentially we completed a study to identify if the AI developed at Originality.AI can detect if the content that was produced by ChatGPT, GPT 3.5 (DaVinci-003) with the same accuracy as it can for GPT-3 which is 94%.
GPTZero is an artificial intelligence (AI) content detection tool created by Princeton University student Edward Tian. It was launched in January 2023, just months after the release of ChatGPT in November 2022. Learn about GPTZero, its 2026 acquisition by Superhuman (Grammarly), and how it works in this GPTZero review.