Copyright

AI-generated video is rapidly gaining traction across the internet, but its legality is becoming a hot topic of discussion. Many creators are leveraging artificial intelligence to produce content that often incorporates copyrighted material, raising questions about intellectual property rights and fair use.

While some experts advocate for viewing these videos as transformative art protected under fair use laws, others argue that using copyrighted material without permission constitutes infringement. This ongoing debate highlights the complexities of copyright law in the digital age and the need for clearer guidelines as technology continues to evolve.

In a significant move that underscores the ongoing tensions between traditional media and emerging technologies, two of Japan’s leading newspapers, Asahi Shimbun and Nikkei Inc., have filed a lawsuit against the US-based generative AI startup, Perplexity AI. The lawsuit claims that Perplexity AI has infringed on copyright by using their articles without proper authorization, raising critical questions about the ethical use of AI in content generation.

The newspapers allege that while Perplexity AI cites them as sources, the information provided by the AI often contains inaccuracies that deviate from the original articles. This not only misrepresents the news but also threatens the credibility of established journalism, which relies on accuracy and trustworthiness. Asahi and Nikkei are seeking an injunction to prevent further use of their content and are demanding damages amounting to approximately 30 million dollars.

In a world where AI-generated art is becoming increasingly prevalent, a recent study has revealed a troubling truth: existing tools designed to protect artists’ work from unauthorized use are not as effective as hoped. Researchers have shown that popular protection tools like Glaze and NightShade can be easily bypassed, leaving creators vulnerable to having their copyrighted material used without consent in AI training datasets.

The study, led by a team of international researchers, introduced LightShed, a new method that can detect and reverse-engineer the distortions applied by these protective tools. With an astonishing accuracy rate of 99.98%, LightShed effectively strips away the protections, rendering the images usable again for generative AI models. This revelation underscores the urgent need for more robust and adaptive defenses in the rapidly evolving landscape of AI and digital creativity.

The recent workshop on “Generative AI and Copyright” held by the Committee on Legal Affairs (JURI) has brought to light a pressing issue: the urgent need for accountability from AI companies that profit from the unauthorized use of creative works. As highlighted by Nina George, the Commissioner for Political Affairs at the European Writers’ Council (EWC), the exploitation of authors’ labor and intellectual property by non-European AI firms raises significant ethical and legal concerns. With millions of protected works being used without consent, the creative community is demanding action.