Google is a ‘bad actor’ says People CEO, accusing the company of stealing content
News Summary
Neil Vogel, CEO of People, Inc., a major U.S. digital and print publisher, has publicly accused Google of being a "bad actor" for using its search crawler to simultaneously scrape publisher content for Google's AI products. Vogel highlighted that Google Search traffic to People Inc. has plummeted from approximately 65% three years ago to the high 20s, yet Google continues to leverage their content for AI without compensation. Vogel emphasized that publishers need more leverage in the AI era, advocating for blocking unpaid AI crawlers. People Inc. has a content deal with OpenAI and is using Cloudflare's solution to block crawlers from other large LLM providers, aiming to force new content deals. However, Google's crawler cannot be blocked independently, as doing so would cut off the approximately 20% of search traffic it still delivers, which Vogel views as intentional "bad actor" behavior. Cloudflare CEO Matthew Prince predicted that by next year, Google will be paying content creators for crawling their content and using it in AI models, suggesting that new regulations might prompt these changes in the future.
Background
The rise of generative Artificial Intelligence (AI) technologies, such as Large Language Models (LLMs), has dramatically increased the demand for training data, including vast amounts of text and image content scraped from the internet. This has ignited heated disputes between content creators and publishers, and AI companies, over copyright, content usage rights, and fair compensation. Many publishers, including People Inc., argue that AI companies are using their copyrighted content without permission or payment to train AI models, thereby undermining their business models and creating direct competitors. This tension has led publishers to explore new strategies, such as blocking AI crawlers or negotiating content licensing agreements, in an effort to secure a fair share in the AI economy.
In-Depth AI Insights
What are the deeper strategic implications of Google's 'single crawler' approach amidst publisher backlash? - Google's 'single crawler' strategy highlights its core dilemma in balancing its traditional search dominance with the development of nascent AI products. It attempts to leverage existing web indexing infrastructure for AI to reduce costs and accelerate AI model training while maintaining its grip on search traffic distribution. - This approach also reflects Google's desire to avoid content licensing costs and its apprehension of a potential collective boycott by publishers against its AI scraping. By tying search traffic to AI crawling, Google gains significant leverage in negotiations with publishers, forcing them into a difficult choice between losing traffic and having their content used without compensation. - However, this strategy may prove unsustainable amidst growing regulatory scrutiny and strong publisher opposition, especially given the Trump administration's critical stance on large technology companies. The Cloudflare CEO's prediction suggests that even within Google, there might be intense debates about changing this strategy, signaling a potential shift towards a content payment model in the future. How might the publisher strategy of blocking AI crawlers fundamentally alter the economics of content creation and AI model training? - This strategy is redefining the power dynamics between content owners and AI developers. By restricting access to high-quality training data, publishers are compelling AI companies to treat content as a valuable asset that requires payment, rather than a freely available public resource. - This will foster a new market for content licensing, creating new revenue streams for publishers and potentially incentivizing investment in high-quality, original content. For AI companies, it means a significant increase in the cost of training models and delivering services, which could impact their profitability or drive them to explore more efficient, data-scarce training methods. - In the long run, this could lead to more diversified data sources for AI model training and encourage AI companies to forge more collaborative relationships with content creators, fostering a fairer and more sustainable digital content ecosystem. What are the long-term regulatory and legal risks for AI companies, particularly given the incumbent US administration's stance on tech power? - AI companies face escalating legal challenges, particularly concerning copyright law, where existing legal frameworks are ambiguous in the context of AI. Recent settlements by companies like Anthropic with publishers suggest that even startups can face substantial liabilities, setting significant precedents for other AI players. - Regulatory risks are on the rise. The US federal government, particularly under President Trump, has been vigilant about large tech companies' use of market dominance and potential anti-competitive practices. This critical stance could spur stricter legislation requiring AI companies to pay for content usage and potentially lead to antitrust actions against those deemed "bad actors." - Future regulations might extend beyond copyright to cover data privacy, algorithmic transparency, and accountability for AI-generated content. These factors could significantly increase the operational complexity and compliance costs for AI companies, potentially impacting their pace of innovation and market valuations.