As Canada and the European Union near a Digital Trade Agreement – News Media Canada CEO Paul Deegan and News Media Europe executive director Wout van Wijk together issue a warning about the threat posed to journalism by unfettered big tech.
Artificial intelligence companies are flagrantly scraping and summarising content directly from published news articles via retrieval-augmented generation. News media is by far the most frequently cited source of current information for LLMs, who are using news content without authorisation or fair compensation. Because AI overviews combine many sources of unlicensed content to provide an effective substitute to the original source. With the user staying within big tech’s increasingly taller walled garden, rather than being pointed electronically to news websites via links, publishers are deprived of audience and their ability to sell advertising and subscriptions is significantly diminished. No clicks mean no cash for news businesses. Yet, AI companies are selling ads against copyrighted (and often paywalled) content as well as subscriptions for their premium products.
AI slop can be harmful to readers
There are plenty of examples of AI overviews serving up slop: inaccurate, irrelevant, out-of-date, and even harmful information. That is because the large language models do not adhere to journalistic standards; they cannot perceive reality, truth, or facts without (mis-)using the work of human journalists and content creators. According to the BBC, “AI assistants have significant issues with basic factual accuracy…The range of errors introduced by AI assistants is wider than just factual inaccuracies. The AI assistants we tested struggled to differentiate between opinion and fact, editorialised, and often failed to include essential context. Even when each statement in a response is accurate, these types of issues can result in responses which are misleading or biased.”
If publishers cannot monetise content, they cannot reinvest in the accurate and authoritative journalism readers rely upon to make informed decisions that empower them to participate effectively in democratic processes.
AI companies are stealing our content and using it to compete against us. They are strip-mining and cannibalising proprietary content and are free riding on the backs of news publishers while unlawfully enriching themselves. In essence, they are unlicensed news syndicators. That’s unfair. That’s anti-competitive. It runs counter to the interests of the news media and the wider public, while undermining government’s drive to encourage AI development and adoption, as this depends on access to high-quality data and information created by humans.
A possible Canada-European Union Digital Trade Agreement could help scale Canadian and European innovation leaders, while ensuring the ethical, positive and responsible use of AI through reasonable guardrails: First, intellectual property should be protected. Second, platforms should provide fair compensation to publishers. Third, platforms should provide clear attribution to source content. Fourth, publishers should be allowed to opt out of AI overviews without their websites being removed from search engines. Fifth, platforms should not discriminate in the ranking of search results.
In negotiating a possible DTA, we recommend trade negotiators consider the Global Principles for Artificial Intelligence, which were developed in 2023 by twenty-six organizations around the world, including News Media Canada and News Media Europe. Those principles include:
- Respecting intellectual property rights, protecting the organizations’ investments in original content
- Leveraging efficient licensing models that can facilitate innovation through training of trustworthy and high-quality AI systems
- Providing granular transparency to allow publishers to enforce their rights where their content is included in training datasets
- Clearly attributing content to the original publishers of the content.
Copyright protections must be properly enforced
We also encourage Canadian and European negotiators to consider the power imbalance between news publishers and dominant online platforms and services, and to put in place measures to address the massive bargaining imbalance.
The US Administration’s AI Action Plan seeks “to achieve and maintain unquestioned and unchallenged global technological dominance”, and it does not even mention the word “copyright”. Against that backdrop, Canada and the EU should align through regulatory cooperation to support homegrown, decentralised, fair and responsible tech development, while protecting intellectual property, so that news publishers can continue to invest in fact-based, fact-checked original high-quality news content produced by real journalists.
Real trustworthy news is an antidote to the proliferation of misinformation online. With a framework backed up by the teeth of enforcement, it contributes to the sustainability of reliable, innovative AI models themselves.
A free, independent, and pluralistic press is a cornerstone of Western democracy. Journalism plays a critical role in holding power to account, informing citizens, and providing resilience against the rising tide of information manipulation, and state-sponsored propaganda.
But today, the unchecked spread of AI-generated content by platforms and chatbots — delivered without fact-checking, transparency, or legal accountability — risks eroding public trust in professional verified editorial sources. When synthetic information is treated as equivalent to journalism, the very foundations of public discourse and democratic decision-making are weakened.
To ensure our free and plural press remains commercially viable, AI providers should not use publishers’ content to build and run their products without consent, credit, and compensation.
Email pged@pressgazette.co.uk to point out mistakes, provide story tips or send in a letter for publication on our "Letters Page" blog
