After nearly 25 years as the founder of Mumsnet, I thought I wouldn’t be shocked at all when it comes to how big tech works. However, last week I read that Google is reviewing UK copyright law, with a view to making it free to mine other publishers’ content for free for commercial gain. That’s when my jaw hit the floor.
At Mumsnet, we have taken a sharp stance against this practice and recently launched the UK’s first legal action against tech giant OpenAI. Earlier this year, we noticed that the virus was scraping content, presumably to train large-scale language models (LLMs). Because such unauthorized scraping is a violation of copyright law and explicitly violates our Terms of Service, we consulted OpenAI and proposed a license agreement. After long negotiations (and nondisclosure agreements signed), the company told us it wasn’t interested, saying it wanted a data source that was “less open.”
You may be wondering why lifting the ban on online content for model training is a problem – Google has been crawling all over websites and ingesting that data for search purposes since the dawn of the internet. Why not? That’s true, but there’s a clear value exchange in giving Google access to that data, the search traffic that comes from being indexed by Google. In contrast, LLM builds models such as ChatGPT that provide answers to every question you can imagine. This means people no longer need to go elsewhere to get solutions. And they build those models using illegally scraped content from the very websites they are trying to replace.
Allowing AI companies to simply steal content is not only grossly unfair to publishers who see no return on the work and risk they put in, but it is also an existential threat to publishers. It’s also threatening (and ultimately counterproductive). If publishers die because AI polices all their traffic, who will be left to create the content that feeds the models? Let’s face it, there’s no way the tech giants can afford to properly compensate publishers. Not. OpenAI has now raised $6.5 billion, the largest single venture capital round in history, and is valued at $150 billion. In fact, it was just reported that the company is planning to change its structure and become a for-profit company.
Some large publishers with legal and financial power have been successful in terminating licensing agreements with AI giants, and several others have filed lawsuits to protect their rights. However, small publishers will be left behind and may never receive compensation if Google and others are at the mercy of copyright laws.
Mumsnet is in a stronger position than other companies to resist the onslaught of AI because much of its traffic comes directly from it rather than through search engines. AI chatbots can spit out ‘Mumsnet-style’ answers to parenting questions, but they’re never funny when it comes to parking spot fights, they’re brutally dishonest when it comes to relationships, and they have about 1,000 It also never provides the emotional support that people see. We estimate that every year, another Mumsnet user contributed to a woman leaving an abusive partner. But if these multi-trillion dollar giants are allowed to play roughshod over content creators and get away with it, they will destroy much of it and all the jobs that depend on them. I’m going to do it.
I’m not anti-AI. It clearly has the potential to advance human progress and improve our lives in countless ways. Mumsnet used this to build MumsGPT. MumsGPT has licensed OpenAI’s API (Application Programming Interface) to uncover and summarize what parents are thinking, from beauty trends to supermarkets to politicians, and build on it. Additionally, we believe there are some good reasons why these AI models need to ingest Mumsnet conversations to train the models. The more than 6 billion words on Mumsnet are a unique record of 24 years of women’s interactions on everything from world politics to relationships with in-laws. In contrast, most content on the web is written by men, for men. AI models have built-in misogyny, and we want to help them counter their gender biases.
But Google’s proposal to change the law would allow multibillion-dollar companies to freely trade around the concept of fair value exchange in the name of rapid “development.” Everything that’s unique and great about small publisher sites will be lost, and a handful of Silicon Valley giants will have even more control over the world’s content and commerce.
It doesn’t have to be this way. There is more than enough money flowing into AI companies for everyone to be rewarded fairly and sustainably for their contributions. But we, the publishing industry and the government, need to wake up and smell the coffee. Because, as the recent Google antitrust case in the United States showed, big tech companies, if left to their own devices, are willing to ignore the law and take tough measures. to increase their advantage.
Justin Roberts is the CEO of Mumsnet
Do you have an opinion on the issues raised in this article? Click here if you would like to email your answer of up to 300 words to be considered for publication in our email section.
