Google is preparing to release a system that can tell if a picture was taken with a camera, altered using Photoshop, or created using generative artificial intelligence models. Google’s “about this image feature” will be modified in the upcoming months to inform users whether an image was produced or altered using artificial intelligence (AI) techniques.
One of the biggest organisations attempting to address AI-generated photography is the Coalition for Content Provenance and Authenticity (C2PA), which includes the system Google is utilising.
The technical standard for C2PA authentication creates a digital trail by integrating information about the origins of images and operating on both hardware and software. Although C2PA authentication has been supported by Amazon, Microsoft, Adobe, Arm, OpenAI, Intel, Truepic, and Google, adoption has been sluggish.
The initiative’s first major test will be the incorporation of Google into search results. Google will utilise the most recent C2PA technical standard (version 2.1), which it contributed to the development of, in conjunction with an upcoming C2PA trust list that will enable platforms such as Google Search to verify the source of material.
According to Laurie Richardson, Google’s vice president of trust and safety: “For example, if the data shows an image was taken by a specific camera model, the trust list helps validate that this piece of information is accurate.”
Google intends to include C2PA metadata in its advertising networks as well, BrandSpur technology and information news reports.
According to Richardson: “Our goal is to ramp this up over time and use C2PA signals to inform how we enforce key policies.”
Adding: “We’re also exploring ways to relay C2PA information to viewers on YouTube when content is captured with a camera, and we’ll have more updates on that later in the year.”
Although Google was among the first major tech businesses to implement the C2PA authentication standard, there will be many adoption and interoperability hurdles before this can be made to function on a wide range of hardware and applications.
The C2PA open technical standard, which adds metadata about camera settings and the location of an image’s capture, is only supported by a small number of Leica and Sony cameras. In addition to Apple and Google pledging to support C2PA on iPhones and Android devices, Nikon and Canon have also committed to adopting the standard.
Affinity Photo, Gimp, and numerous other programs do not support adding C2PA data, while Adobe’s Photoshop and Lightroom can. The majority of large internet sites do not provide labels, thus there are further issues with viewing the data once it has been placed to a photo. Still, the way Google is used in search results might inspire other search engines to use comparable labels.
Continuing, Richardson admitted: “Establishing and signaling content provenance remains a complex challenge, with a range of considerations based on the product or service.”
“And while we know there’s no silver bullet solution for all content online, working with others in the industry is critical to create sustainable and interoperable solutions,” he added.
