Why Meta is paying for controversial content – and what that means for the posts you see

In recent days, media reports have revealed several instances where Facebook has paid for content that could be deemed controversial.

ABC News Verify found that Facebook had monetised Australian accounts described as having white nationalist, neo-Nazi or anti-vaccine links. According to the report, some of the content they posted even appeared to conflict with Meta’s own monetisation policies.

Meta did not answer questions about the individual pages, and their earnings are not public. However, it did state that if creators breach their policies, their earnings will be disabled, and they may be removed entirely for repeated breaches.

Payment is different from simply allowing contentious speech to remain online. It creates a commercial relationship between platforms and the people who post content on them, and gives creators a reason to produce more of what performs.

“Meme factories” in countries like Indonesia, India and China exploit this, posting content on pages such as pro-One Nation groups without really knowing who Pauline Hanson is.

So how does it work, and what does it mean for the content you see?

How Facebook content monetisation works

Meta introduced its Facebook content monetisation program in 2024 by combining several previous monetisation features, such as ads on reels. Eligible creators can earn money from public videos, stories, photos and text posts.

The program remains invitation-only with no publicly available criteria on whom Meta will invite. However, creators can register their interest to join.

Meta describes the system as performance-based. Essentially, the more engagement your content receives, the more money Meta will pay you. In 2025, Facebook paid creators nearly US$3 billion.

In March this year, the company said payments would place greater emphasis on “qualified views” (views Meta deems eligible), longer watch time and deeper engagement.

Creators can see an approximate earnings rate per 1,000 qualified views, but Meta doesn’t publicly disclose how it weights those signals or calculates individual payments.

There are three layers here that we should address separately. Community standards determine whether material may remain on Facebook. Distribution and recommendation systems influence how widely it travels. Monetisation rules determine whether eligible content earns money.

Meta acknowledges these layers. Content can be permitted to be on the platform, but excluded from recommendations. Material that doesn’t breach Meta’s community standards can still be distributed less widely – for example, if Meta thinks it’s low quality. And content that’s otherwise fine may still not be monetised, because the creator isn’t part of the program.

In short, content remaining online, being widely circulated, and being paid for are not the same thing on Facebook.

Why inflammatory content can still earn

While Meta does have rules governing the content on Facebook, the problem is that separate systems for ranking, recommendation, payment and enforcement might not operate consistently at scale.

In my work on digital media ecologies, I use “medialogy” to examine digital systems as connected arrangements of interfaces, data, institutions, commercial interests and everyday practices. This moves the debate beyond a vague claim that “an algorithm” caused the problem.

Facebook’s creator economy datafies attention: it turns viewing, watching and engaging into measurements that organise visibility and economic value. Creators use this. They study dashboards and audience responses, learning which subjects, formats and tones travel furthest.

Research suggests morally charged material can benefit from these dynamics. A large study last year found moral-emotional language was positively associated with sharing across multiple topics and datasets, although effects varied. Another study found misinformation evoked more outrage than trustworthy reporting, and that outrage facilitated sharing.

This doesn’t mean every angry post succeeds, or that recommendation systems alone cause polarisation. Exposure to misinformation and extremist material is often concentrated among relatively small groups, and audience demand also matters.

When payment depends on performance, and provocative material performs well against relevant measures, harmful content can become financially valuable – unless the platform enforces policies that reliably counteract that incentive.

The evidence doesn’t show Meta intends to fund extremism. It shows its performance economy and safety regime can work at cross purposes.

Are other platforms doing the same?

Not in precisely the same way.

YouTube shares advertising and subscription revenue. Its terms allocate creators 55% of net watch-page advertising revenue and 45% of shorts revenue. Separate advertiser-friendly rules can limit earnings from hateful, incendiary or demeaning content, even when a video remains online.

TikTok’s creator rewards program, where available, pays eligible videos after 1,000 qualified views in the For You feed. Its formula considers watch time and completion, search traffic, engagement, location and advertising value. Eligible videos must be original and at least one minute long.

X (formerly Twitter) is currently changing its monetisation program. From September 8, eligible creators will be able to earn revenue for impressions their posts get on the home feeds of premium users. Misleading content, adult content and hate or extremist material may be ineligible or restricted.

What does monetisation mean for what we see online?

The recurring risk is similar among platforms, regardless of the specific monetisation program. Platforms translate measurable attention into money, then rely on separate policy systems to prevent harmful material from benefiting.

Accountability requires public disclosure over which account and content categories receive payments. There should be independent audits connecting recommendation and monetisation decisions. Researchers should have privacy-protected access to distribution and earnings data. And repeated breaches should trigger timely, reviewable demonetisation.

The question is not whether platforms have rules. It’s whether they can demonstrate that their payment systems do not undermine them.