The European Commission has initiated an antitrust investigation into global tech giant Google for allegedly flouting European Union (EU) competition rules by utilising web publishers’ content and YouTube videos to train its artificial intelligence (AI) models, a press release dated December 9 reveals. The EU’s primary executive branch says the investigation will centre on whether Google is imposing unfair terms on independent web publishers and content creators, and consequently placing rival AI model developers at a disadvantage. Notably, the present antitrust investigation stems from a complaint filed by a group of independent publishers with the European Commission in July this year. For context, these publishers accused Google of misusing its market power within the online search domain, which, in turn, hurt their web traffic numbers. “Google’s core search engine service is misusing web content for Google’s AI Overviews in Google Search, which have caused, and continue to cause, significant harm to publishers, including news publishers, in the form of traffic, readership, and revenue loss,” the publishers’ alliance stated in its complaint. What Is The European Commission Investigating Exactly? With respect to the current antitrust investigation, the European Commission says it is concerned that Google may have: 1) Used web publishers’ content to power its generative AI services, namely AI Overviews and AI Mode—on its search results page without providing commensurate compensation to publishers. The Commission is also concerned that Google did this without offering publishers the opportunity to refuse such use in the first place. As such, the EU’s executive arm says it will investigate the extent to which results displayed as part of AI Overviews and AI Mode are based on web publishers’ content, and whether publishers can refuse such use without losing access to Google Search. 2) Used content on YouTube to train AI models without commensurate compensation to content creators, and without giving them the opportunity to refuse such use. For context, content creators uploading videos on YouTube are obligated to grant Google permission to use their data for different purposes, including training generative AI models. Interestingly, Google does not remunerate YouTube content creators for this use, nor does it allow them to upload content without granting Google the right to utilise such data. At the same time, rival AI model developers are barred from using YouTube content to train their own models. “AI is bringing remarkable innovation and many benefits for people and businesses across Europe, but this progress cannot come at the expense of the principles at the heart of our societies. This is why we are investigating whether Google may have imposed unfair terms and conditions on publishers and content creators, while placing rival AI models developers at a disadvantage,” EU antitrust chief Teresa Ribera remarked. What Does This Investigation Mean From India’s Standpoint? From India’s standpoint, this development is crucial, as the country is YouTube’s largest market . As such, Google using content uploaded on the platform to train its AI models becomes a major data governance concern. Notably, this alleged use might be happening without Indian YouTubers receiving any proportionate compensation or the opportunity to refuse such data use in the first place. All content hosted on YouTube is licensed by the respective creators. However, Google requires them to “grant certain rights to YouTube”. The video platform’s terms specify that it may use an uploaded video to “reproduce, distribute, prepare derivative works, display and perform it”. YouTube may also use that content in connection with its own or its affiliates’ business activities. However, the terms of service do not explicitly mention training AI models, or using video content to enhance features such as AI Overviews or Google Gemini. Advertisements Interestingly, Google gives content creators on YouTube the choice to ‘allow third-party training’ of AI models. To explain, creators and rights holders can allow ‘third-party companies’ to use their YouTube videos in order to train their AI models. To do this, they can switch on the third-party training setting, and select which companies to allow. Notably, third-party AI training is only possible for YouTube videos: whose privacy setting is public that fall in line with YouTube’s terms of service and community guidelines Google’s AI Tools Suite On YouTube And Their Implications Google has already rolled out an “AI-powered search results carousel” on YouTube. For example, when someone searches for topics like shopping or travel, the feature presents an AI-generated carousel showing relevant clips from a range of YouTube videos. YouTube also has a conversational AI tool that answers users’ questions, suggests relevant content, and offers AI-generated comment summaries beneath videos. Google claims that the latter feature organises comment sections of long-form videos into digestible themes, helping people understand and participate in conversations more easily. Creators can also use these summaries to participate quickly in comment discussions and draw inspiration for new content. Importantly, while these features appear beneficial on the surface, they raise legitimate concerns about how Google and YouTube are using videos and comments to build and refine their own AI systems. What Can the CCI Do Here? In light of the European Commission’s actions, the Competition Commission of India (CCI) should also consider launching an investigation into how YouTube is using uploaded videos in India to train and refine its AI models. This is especially pertinent as YouTube’s terms of service do not mention AI or permissions related specifically to using uploaded content for AI training. Further, the CCI can take a leaf out of the Department for Promotion of Industry and Internal Trade’s (DPIIT) book and regulate through a framework where AI developers — Google in this case —obtain lawful access to copyrighted material and use it for AI training without negotiating individual permissions or licensing contracts. Elsewhere, the CCI recently proposed a framework that requires companies using AI systems to conduct internal audits of their algorithms and identify potential anti-competitive effects. In this context, the CCI must also demand that Google conduct such audits proactively and curb any anti-competitive practices relative to other AI developers, if such issues are identified. Also Read: DPIIT Committee Proposes Mandatory AI Licensing Model With Royalties And No Opt-Out For Creators Why CCI Wants Indian Firms to Self-Audit AI Systems for Competition Risks How India’s Competition Regulator Plans to Tackle AI-Driven Market Concentration
TechnologyTechnology
As EU Probes Google for Training AI With Creator Content, What Should CCI Do?
Loading article...
AI Index
Neutral / Balanced
Facts presented without strong bias.