2025 was the year AI slop went mainstream. Is the internet ready to grow up now?

From aggressive AI chatbots to nonsensical e-commerce product summaries, “AI slop” really hit its stride in 2025, polluting search engines, shopping platforms, and even the White House’s official communications channels.

The term, which defines low-quality and generally unwanted AI-generated content, has been circulating online since the early 2020s, but data shows it reached a new peak this year.

Online media, social, and consumer intelligence company Meltwater found that in 2025, mentions of “AI slop” across the Internet increased ninefold from 2024, with negative sentiment hitting a high of 54 per cent in October.

The slop (and discourse around it) won’t stop – AI-generated articles now make up more than half of all English-language content on the web, according to search engine optimisation (SEO) firm Graphite.

AI slop even earned the 2025 Word of the Year title from Merriam-Webster and Australia’s national dictionary.

As 2025 draws to a close, we look back at how AI features have changed some of the most-used sites on the internet, and share predictions for what 2026 might look like on the World Wide Web.

‘Solutions looking for a problem’

Amid the rise in AI slop, the ongoing AI hype fuelling financial markets, and a race to dominate the new economy emerging around these technologies, product designers have found themselves stuck in the middle, according to some analysts of user experience (UX) design.

“In the design space, there's a lot of pressure to show the shareholders, ‘Look, we put AI in our product,’” said Kate Moran, vice president of research and content at Nielsen Norman Group, a research and design firm that advises global researchers and user experience (UX) designers.

“This is technology-led design, starting with the tool, and then trying to look for a problem that potentially that tool could solve. There's this pressure to start from the solution and work backwards to find the problem, which is antithetical to how design should be done,” she told Euronews Next.

In past years, product designers have been tasked with integrating AI anywhere and everywhere, Moran said, even when it might not make sense.

She gave the example of Meta, which introduced an AI search function on Instagram last year, replacing the traditional search bar.

“They backpedalled really fast because I’m sure people were furious,” she said. “You believe a search bar does a certain thing, and then all of a sudden, when you start typing in there, you’re talking to an AI chatbot and you didn’t want that. That’s a bad experience.”

This year, AI slop moved into the physical world as well, with AI-focused consumer hardware, such as Humane AI Pin, receiving scathing reviews from users and executives – including Logitech’s CEO Hanneke Faber.

"What's out there is a solution looking for a problem that doesn't exist," Faber said in an interview with Bloomberg.

From active backlash to passive disinterest

Meta has been particularly active in embracing AI tools and AI-generated content, launching a new app dedicated entirely to “AI slop” this year.

In response to OpenAI’s Sora app, which helps users generate videos with AI, Meta introduced “Vibes” to European markets in November. The company describes the platform as “a brand new feed where you can create and share short-form, AI-generated videos, remix content from others, and explore a world of imaginative possibilities.”

But according to internal data seen by Business Insider, Vibes hasn’t made much of a splash in Europe – bringing in just 23,000 daily active users in the first weeks after its launch. The largest audiences were found in France, Italy, and Spain, according to the report, with 4,000 to 5,000 daily active users in each country.

The company doesn’t publish user figures publicly, and did not respond to Euronews Next’s request for comment.

Meta’s pivot to AI-generated content is particularly perplexing given that earlier this year, the company said it was tackling “unoriginal content” and advising creators to favour “authentic storytelling” over short, low-value videos.

Why more AI isn’t always the answer

According to Moran, flashy AI tools often aren’t what users are looking for in a digital product or service. They can confuse users by challenging their basic understanding of how these services are meant to work in the first place, what she calls their “mental model”.

For Daniel Mügge, a researcher at the University of Amsterdam who studies the European governance of AI through the RegulAite project, this represents a bigger issue with tech companies’ priorities.

“I think what is clear and should be genuinely worrying is that a number of these companies have been engaged in a kind of race among themselves,” he told Euronews Next.

“(It seems there’s) a bit of an all-or-nothing desperate gamble for other large tech companies that are basically betting the house on whether they can still find a way of beating OpenAI at their own game,” he added.

Generative AI has especially received a disproportionate amount of attention, according to Mügge, considering its relatively limited economy-wide impact. He said the European Union would be better off investing in AI that tackles specific societal issues, such as robotics or manufacturing.

“We see that quite a bit of AI investment actually ends up in applications that make society a worse and not a better place,” Mügge said, citing AI tools in advertising as an example of an application he considers actively harmful.

“That's the sort of investment that I think we actually don't need, and if we don't have it here in Europe, for example, it's a good thing rather than a bad thing.”

Could ‘boring’ AI be better in the long run?

Backlash to the amount of AI slop taking over the internet has led some sites – including Pinterest and YouTube – to introduce features that allow users to limit the amount of AI-generated content they see.

Features like these, along with more practical AI tools that require less user interaction, are bright spots in online user experience, Moran said.

“The things that this technology can do that are really useful, that I think actually are changing the products that we design and the way that we work, are not the sexy things,” she said.

She cited Amazon’s AI-generated summary of product reviews as a useful, if boring, AI feature, which enhances the user’s experience without fundamentally changing how they interact with the service.

“Being able to give a quick qualitative summary of how people feel about that product is really valuable and it requires zero interaction. All people have to do is read it,” she said.

Making room for smaller companies that make useful products which may not grab as much attention could also be a good way for tech ecosystems in Europe to carve out a separate path, according to Mügge.

“I think there's a lot of scope for relatively smaller, much more specialised companies to play a meaningful role in that, and then you don't need to worry so much that there is no European competition to OpenAI,” Mügge said.

Both Mügge and Moran agree that the tide seems to be turning, as AI hype makes way for more intentional product design and strategy that focuses on impact.

“Nobody knows what's going to be next or where the technology is going to evolve from here,” Moran said.

“Right now, these smaller, more narrowly-scoped features are a lot easier for people to use, and even if they're not flashy or sexy, they can make a really big difference in people's lives,” she added.