Australian AI company Sapia.ai has launched the world’s first proprietary model to detect AI-generated content in online chat interviews.
Sapia.ai, the world’s only intelligent chat platform powered by deep learning AI, has released a new capability to detect and flag responses by generative AI such as ChatGPT in real time.
This new feature is powered by a proprietary dataset of over 1 billion words that matches over 12 million responses from 2.5 million candidates using Sapia.ai’s platform.
Conversational natural language processing (NLP) based chat AI interviewing assesses and screens top talent at scale via an easy-to-use messaging platform.
In addition to improving diversity outcomes by eliminating unconscious bias, companies can reallocate the thousands of hours spent screening talent to higher value work.
This new feature will prevent candidates from using generative AI tools to respond to prompts from the Sapia.ai platform.
Candidates receive real-time alerts as they answer if their answer is likely to be AI-generated content (AGC), so they can make changes to their answer before final submission.
Failing this, notify the decision maker that the candidate’s response may include AGC and ask for further consideration.
Barb Hyman, CEO and Founder of Sapia.ai, said: It’s our competitive moat. ”
“It is possible to detect generative ChatGPT usage through analytics, but we do it in real time. Our dataset makes it easier to adapt to new iterations of generative AI.” .”
“This is based on the plagiarism flag we announced in 2019. We identified candidates who got their answers from the internet or other candidates. It preserves the integrity of structured interviews as a fair and accurate method of assessment.”
“Equally important to the integrity of our product is being transparent to individuals from the start that their interview responses are being evaluated for AGC and providing them with the option to change their responses in real time. It is that
“We have always built our products with the human experience first. , we aim to earn your trust in that experience.”
“We believe every candidate matters. Today’s applicants are tomorrow’s customers. That’s why we strive for exceptional candidate satisfaction metrics, with an average completion rate of 91%,” says Hyman.
Dr. Buddhi Jayatilleke, Chief Data Scientist at Sapia.ai adds: “We tested his AGC flag using thousands of responses generated from GPT-2, GPT-3 and ChatGPT for various prompts associated with multiple role families.”
“We were able to achieve an ROC-AUC of over 95%, which is a strong indicator of the classifier’s accuracy.”
“This is due to our ability to leverage large human-written response datasets to distinguish between the boilerplate nature of content from human-written text and generative AI models,” said Dr. Jayatilleke. say.
Global brands such as Starbucks Australia, Holland & Barrett and WOLT use Sapia.ai to accelerate and enhance their hiring and promotion processes.
