As 404 Media reported on Februarz 27, 2024, both Tumblr and WordPress.com (both owned by Automattic) are preparing to sell user data to Midjourney an OpenAI, as the team at 404 Media obtained internal documents showing that Tumblr staff compiled users' data as part of a deal with Midjourney and OpenAI. According to the documents obtained by 404 Media, it is unclear what the exact types of data from each platform going to be, but internal communications made it clear that deals between Automattic and OpenAI and Midjourney are imminent.

The post revealed that in a messy and unclear process, Automattic compiled data that it should not share, but it was also unclear if and what has already been shared with OpenAI. For websites and blogs hosted on WordPress.com, there will be an opt-out setting to prevent content getting shared with OpenAI. Yes, you have to opt-out, since obviously no one would opt-in a setting to allow AI scraping your data. WordPress is also offering an open source CMS which people can use for their self-hosted websites. As 404 Media reported, it is unclear "whether self-hosted WordPress websites that use Automattic plugins like JetPack to connect websites with Automattic's infrastructure are subject to the company's AI-scraping deals". I highly recommend reading the superior reporting by Sam Cole from 404 Media to get all the insights on the whole situation.

I am not an opponent of AI, I believe that the technology can be used for good, ethical, and powerful flows and causes, but I am definitely an opponent of setting up such AI-scraping deals and then telling your users to opt-out instead of opt-in. For the longest time, I had the impression that Automattic is "one of the good ones". I built my very first website with WordPress, and I was always fine to pay some money here and there for a additional functionality, hosting, or anything similar. I never had the impression that WordPress, but especially Automattic turns into such an unethical, creepy, privacy-invading company.

According to Web Technology Surveys, 43.2% of all websites online are running on WordPress, which is insane. Imaging that probably a chunk of that number will get used to scrape data for AI must be worth such a huge amount of money, that Automattic just said "go for it". I assume that most people who are using WordPress put their trust into their services, especially since they are one of the biggest players in the open-source field. The people who appreciate the ease and quickness of setting up a WordPress website to just write and publish a blog suddenly have to find out that the company powering their websites became a victim of the disturbing, data-scraping field AI has become recently. WordPress has opened the door to the internet for so many people, but it seems like that it is finally time to ditch it.

By agreeing or factoring in such deals, Automattic has shown its true colors. While, as of now we are only talking about Tumblr and websites hosted on WordPress.com, but there are already concerns about self-hosted WordPress websites which are using Automattic's JetPack plugin, it is safe to say that you better migrate off WordPress for the good, simply to be not exposed to the risk of getting your data scraped. What is equally worying and yet another alarming part of that deal is the fact that Automattic not only owns Wordpress.com and Tumblr but also apps like Pocket Casts (a Podcast Player), Simplenote, a cross-platform, simple, and free note-taking app that has often been praised for its privacy focus, as well as Day One, which is probably the most popular journaling app used by people to input their notes, thoughts, ideas, images, and more, so basically it is a gold mine for data. I am not saying that Automattic will expand its deals across those products too, but simply the fact that they own them, knowing that they are fine scraping and selling their users' data is indeed worrying. With this AI-scraping deal, Automattic showed its real face, it showed what they have planned, how they will make more money, and that they do not care about their users.

As reported by Sam Cole within the 404 Media post, Automattic shared content posted on Tumblr and WordPress between 2014 and 2023 without users' permission, beyond that, the data that has been exchanged included private and deleted posts, private answers, and content from premium partner blogs. How can you trust a company who did that long-term? The simple answer is, you can not, you can not trust them, and if you are using Tumblr or WordPress, it is the right tome to seriously consider moving your website to an CMS alternative. Speaking of that, Creativerly runs on Ghost* since years, and while it has been the most joyful experience for me personally, there are also other lovely WordPress alternatives like Kirby, Craft, or simply blogging and writing tools like Write.as. Besides that, I created a blog post which contains a bunch of website builders, blogging platforms, and CMS to quickly and easily build your personal website, blog, or publication.

WordPress has once been the home for a lot of writers and publications. Those are the folks who turned WordPress into what it is today. Automattic treated that situation with the outmost disrespect.

It is time to finally ditch WordPress.