OpenAI To Introduce New Tool To Detect AI-Generated Images; All You Need To Know

ChatGPT maker OpenAI will roll out a new tool to spot images made by its DALL-E 3 text-to-image generator. The AI company said that the tool, tested internally, accurately pinpoints DALL-E 3 creations about 98% of the time, even handling tweaks like compression and cropping well.

With crucial elections taking place across the globe, the move comes with widespread concerns about AI-generated content being used to spread misinformation.

The company said that starting Wednesday, it is opening applications for access to OpenAI's image detection classifier to their first group of testers - including research labs and research-oriented journalism nonprofits - for feedback through their Researcher Access Program.

"Our goal is to enable independent research that assesses the classifier's effectiveness, analyzes its real-world application, surfaces relevant considerations for such use, and explores the characteristics of AI-generated content," the company said.

The company is also teaming up with major players like Microsoft to bolster content authenticity standards. This includes joining the Coalition for Content Provenance and Authenticity (C2PA), ensuring digital content can be traced back to its source.

Why the fuss? Fake videos circulating during India's ongoing elections featuring Bollywood stars criticising Prime Minister Narendra Modi highlight the growing use of AI-generated content in influencing voters.

OpenAI's efforts extend beyond just spotting AI-made content. They're developing tamper-resistant watermarking and detection tools to safeguard digital integrity. The watermarking adds invisible markers to content, making it tough to alter, while the detection tools use AI to flag potentially AI-generated content.

Through a $2 million societal resilience fund with Microsoft, OpenAI aims to boost AI education, preparing people to navigate the evolving digital landscape.