This year, the percentage of ad spend going toward made-for-advertising (MFA) sites went up instead of down for the first time since 2023.
That’s according to the Association of National Advertisers’ (ANA) programmatic benchmark report for Q1, released in May.
It’s a notable reversal, all things considered. The ANA first kicked off an industrywide anti-MFA crackdown with its 2023 report, which revealed its members spent 21% of their budgets – about $13 billion annually – on MFA. Since that report, and following countless attempts by programmatic ad tech companies to clean up their supply chains, MFA spending has been consistently trending downward – until Q1, that is.
Granted, the percentage of ANA members’ ad spend flowing to MFA properties only increased from 0.6% in Q4 2025 to 1.1% in Q1 2026. So MFA is still getting just a small slice of ad budgets, even if its share has recently doubled in size.
But that statistic doesn’t tell the whole story, said Julie Weitzner, the ANA’s SVP of media practice. Spending on MFA sites is concentrated among advertisers whose campaigns don’t perform as well as their peers’. The lower-performing cohort of ANA advertisers spent 2.1% of their budgets on MFA, compared to 0.9% for the highest-performing group.
Plus, the ANA’s report hints at a potential culprit for MFA’s slight resurgence. The increase, it reads, serves as “a reminder that MFA requires ongoing mitigation, particularly with growing sub-types such as AI slop.”
MFA vs. AI
There’s an emerging consensus that AI slop is a growing media quality challenge for advertisers. Critics say generative AI makes it easy for MFA publishers to create and monetize low-effort content across text, audio, video and social media.
“AI slop” usually refers to content generated without human input that’s of dubious news or entertainment value. However, there are conflicting reports regarding just how much of the internet is currently made up of AI-generated content, let alone that which could be defined as “slop.”
Research from AI marketing platform Ahrefs suggests that 74% of new websites created in April 2025 included AI content, and only 26% were “pure human.” Another study from SEO firm Graphite claims that, as of October 2025, 50% of online articles were generated by AI.
That’s a lot of AI content flooding the web in a very short amount of time. Remember, ChatGPT didn’t become available to the public until February 2023.
Meanwhile, MFA and AI slop share a lot of similarities, according to Scott Pierce, senior director of product management and marketplace quality at The Trade Desk, who wrote a column for AdExchanger in November titled, “AI Slop Is The New MFA, And We All Need To Fight It.”
However, even if there are some similarities, AI slop is not the same exact thing as MFA, Pierce said while speaking to AdExchanger for this story.
Most experts agree that, for a website to be considered made for advertising (or “made for arbitrage,” as some prefer), it requires a high ad-to-content ratio, poor-quality content and an overreliance on paid traffic sources. AI slop sites don’t always fulfill all three requirements, although it can happen. More often, AI slop only fits two criteria – poor content and a high degree of ad clutter.
But even if AI slop and MFA aren’t completely the same thing, the methods the industry has devised for demonetizing MFA can also help it avoid AI slop.
For instance, the anti-MFA crackdown has placed tremendous pressure on SSPs to proactively weed MFA inventory out of their supply chains. Which is making it harder for AI slop to monetize, said Rocky Moss, co-founder and CEO ad DeepSee.io, whose supply chain research underpins many companies’ efforts to combat MFA.
One example of how SSPs have become more effective gatekeepers is that they typically won’t include new sites in their bidstream until they’ve been around for a certain period of time, Moss said. This helps prevent MFA publishers from simply spinning up new sites once their existing sites are blocked by SSPs.
But poor-quality AI sites tend to be a lot more ephemeral than even your traditional MFA sites, Moss said, often only lasting online for 30 or 60 days. Which means they don’t stick around long enough to get included in SSP auctions.
Are you afraid of AI slop?
However, the bigger concern when it comes to AI slop sites isn’t the creation of new sites, Moss said. He’s much more worried about older, more reputable publications that are struggling to monetize getting bought and turned into low-quality content farms.
Another challenge is that so many reputable publishers are experimenting with AI technology in their workflows. So how do you draw the line between AI slop and more valuable AI-generated content when even the big publishing brands are involved?
“We make a huge mistake as an industry if we start to say that all AI-generated content is bad,” said Chris Kane, founder of Jounce Media, another company whose research has been instrumental in combatting MFA. “There’s a difference between using generative AI technologies to augment the productivity or the research process versus creating a spam factory, which is automatically publishing hundreds of articles every day.”
From the SSP perspective, however, it’s not worth splitting hairs. “For us, it does bleed together quite a bit from a quality standard enforcement standpoint,” said Jade Grodesky, director of exchange quality at SSP Index Exchange.
Put another way: “The definition of what’s MFA versus slop almost doesn’t matter, because we’re going to take action against both,” said Ilana Wollin, SVP of business ops at Nexxen, whose end-to-end platform includes an SSP and DSP.
Both Nexxen and Index Exchange were proactive about removing MFA from their bidstreams even before it became a hot topic in Ad Land. Index Exchange started its own investigations in 2022 and claimed to be fully MFA-free by 2024. And Nexxen began officially targeting MFA on both the SSP and DSP side in September 2023, citing Jounce’s MFA blocklist as a major resource for those efforts.
Nowadays, Index Exchange uses a generalized exclusion list to keep low-quality content and MFA out of its bidstream, while Nexxen prefers only working with a list of approved publishers. Both rely on other methods as well, including a combination of human-led audits (albeit with the help of some automated tools) that check against data from third-party vendors like Jounce and DeepSee, as well as their own criteria.
If you stay ready …
Regardless of what tools the ad tech industry is using to keep both MFA and AI slop at bay, everyone that AdExchanger spoke to acknowledged that keeping abreast of the latest trends in spam and scammery is a constant battle. And AI is of particular concern because it allows bad faith actors to scale up quickly and cheaply.
“Money follows opportunity,” said Weitzner. “And if there are advertisers that are willing, even at a very small percentage, to put ads up on these sites, it’s worth it to spend next to nothing to create these domains that will potentially garner some ad dollars.”
Ultimately, however, it’s up to individual brands and advertisers to decide where their priorities lie. Are they focused on low CPMs that can bring in cheaper impressions, or are they willing to spend more on higher-quality content that performs better? Because, as the ANA’s most recent report suggests, those MFA ad impressions may not be worth the lower price tag in terms of performance.
“At the end of the day, the focus should be on performance,” said The Trade Desk’s Pierce. “As an industry, if we just turn our focus to meaningful outcomes, we will naturally navigate away from a lot of this.”
