OpenAI, a pioneer in the field of generative AI, is stepping up to the challenge of detecting deepfake imagery amid a rising prevalence of misleading content spreading on social media. The organization behind ChatGPT and the image generator Dall-E is currently testing a tool with an outstanding accuracy of 99% in detecting AI-generated images.
Mira Murati, the brain behind the technology at OpenAI, recently disclosed the effectiveness of the tool at the Wall Street Journal’s Tech Live conference in California. She mentioned that the tool is undergoing internal testing before a planned public release, without specifying a timeline.
With the evolution of technology, AI is now capable of generating everything from images of Pope Francis wearing a Balenciaga coat to harmful fraudulent images targeting users. Exploiting this, an increasing number of scams leveraging AI technology are emerging. As these tools become more sophisticated, distinguishing between what is real and what is AI-generated is becoming a challenge.
In response to this reality, Adobe Inc.’s image generator Firefly addresses another aspect of the challenge by promising not to create content that violates the intellectual property rights of creators.
Not staying behind in the game, OpenAI’s executives have hinted at the AI model following GPT-4. While OpenAI has not publicly disclosed what the next model after GPT-4 might be called, the startup filed a trademark application for “GPT-5” with the United States Patent and Trademark Office last July.
Chatbots like ChatGPT, using GPT-4 and the previous model GPT-3.5, tend to produce inaccurate or ‘hallucinated’ output. When asked if GPT-5 would still face such issues, Murati expressed optimism about future developments.
Sam Altman, the CEO of OpenAI, delved into potential strategic changes. Although currently relying on external giants like Nvidia Corp for chip demands, Altman did not rule out the idea of OpenAI entering the chip manufacturing field, though he mentioned he is currently not inclined towards it.
As OpenAI continues its pioneering journey, the Asia-Pacific region will closely observe and recognize the profound significance of these advancements for the integrity and reliability of digital content.