Photo was created by Webthat using MidJourney
Concerns over Copyright and Transparency
A consortium of media organizations has jointly issued a call for the establishment of regulations safeguarding copyright when utilizing data to train generative AI models. In an open letter addressed to lawmakers worldwide, they advocate for increased transparency in training datasets and the consent of rights holders before incorporating data for AI training purposes.
Negotiating Consent and Combating Bias
The letter also urges authorities to enable media companies to engage in negotiations with AI model operators. The goal is to identify and distinguish AI-generated content while promoting responsible AI usage. Another key point raised in the letter is the demand for AI companies to actively combat bias and misinformation within their services for training data.
Notable Signatories and Their Concerns
Prominent signatories of the open letter include esteemed organizations such as Agence France-Presse, the European Pressphoto Agency, the European Publishers’ Council, Gannett, Getty Images, and more. The collective concern revolves around the dissemination of information by foundation models, which fails to acknowledge original creators and can adversely impact media industry business models.
Impact on Media Industry and Information Access
The signatories emphasize that this practice undermines core revenue streams for media companies, including subscriptions, licensing, and advertising. Beyond violating copyright laws, the consequences extend to reduced media diversity and weakened financial sustainability, ultimately limiting the public’s access to reliable and high-quality information.
Google’s Genesis AI Tool and Errors in AI-Generated Content
Recent developments highlight the issue, with Google showcasing its generative AI news writing tool, Genesis, to several renowned news outlets. Instances of AI-generated articles containing errors further fuel concerns about the reliability of AI-generated content and its implications for the news industry.
Legal Landscape and Collaborative Solutions
The legal status of AI models trained on copyrighted material remains a contentious subject. The matter has attracted attention, including Senate hearings and ongoing lawsuits related to copyright infringement. Amidst these challenges, the signatories of the open letter acknowledge the potential benefits of generative AI and express a desire to contribute to discussions on balancing AI advancements with media companies’ rights.