The ongoing legal saga between Getty Images and an AI startup has taken a significant turn. Recently, Getty Images decided to retract its primary copyright infringement allegations against the AI company, which has been at the center of a high-profile legal dispute regarding the use of copyrighted materials for training artificial intelligence models.
While this development does not conclude the legal proceedings entirely, as Getty continues to pursue other claims and a separate lawsuit in the United States, it highlights the complex and often ambiguous landscape of content ownership in the era of generative AI. This shift in strategy comes on the heels of a recent ruling in the U.S. that favored another AI company in a similar copyright dispute, further complicating the legal framework surrounding AI training practices.
Getty Images initiated legal action against the AI startup in January 2023, accusing it of utilizing millions of copyrighted images without authorization to develop its AI model. The company argued that many of the images produced by the AI were strikingly similar to the copyrighted works used during its training phase, with some even bearing Getty’s watermarks.
As of the latest updates, Getty has dropped these specific claims, signaling a potential shift in their legal strategy. Legal experts suggest that the decision to withdraw the training-related allegations may stem from Getty’s inability to establish a strong enough link between the alleged infringements and the jurisdiction of UK copyright law. Furthermore, the claims regarding the output of the AI model were likely abandoned due to insufficient evidence demonstrating that the generated content significantly reflected the original copyrighted images.
In their closing arguments, Getty’s legal team indicated that the retraction was a tactical move, allowing them to concentrate on what they perceive as more robust claims that could lead to a favorable outcome. This strategic pivot underscores the challenges faced by companies in navigating the evolving legal landscape surrounding AI and copyright.
What remains in Getty’s legal arsenal includes a secondary infringement claim and allegations of trademark infringement. The secondary infringement claim posits that the AI models themselves could potentially violate copyright laws, suggesting that utilizing these models in the UK might be akin to importing infringing materials, regardless of where the training occurred.
A representative from the AI startup expressed satisfaction with Getty’s decision to drop several claims, emphasizing their confidence that the remaining trademark allegations would not hold up in court, as consumers do not interpret the watermarks as commercial endorsements from the AI company.
In addition to the UK lawsuit, Getty’s U.S. division has also filed a separate case against the AI startup, alleging that it used up to 12 million copyrighted images without permission. The damages sought in this case are substantial, with Getty claiming $150,000 for each of the 11,383 works involved, totaling an eye-watering $1.7 billion.
Moreover, the AI startup is also facing legal challenges from a group of visual artists who have filed a copyright infringement complaint against it, along with other AI companies, further complicating the legal landscape for generative AI technologies.
Interestingly, Getty Images has also ventured into the generative AI space, launching its own service that utilizes AI models trained on its extensive stock photography and video libraries, allowing users to create new, licensable images and artwork.