While X charging users $1/year to combat bots and spam is an appealing idea, the value in manipulating X is so high that bad actors will still find a way
X’s Not A Bot trial put a small annual payment in front of posting and interactions for new, unverified accounts in two markets. Mullenweg’s critique focuses on the mismatch between that low barrier and the potentially high payoff from influencing activity on the platform.
The test also sits alongside X’s move toward multiple paid service tiers, making the anti-spam charge part of a broader effort to segment access and features through subscriptions.
First-order effects
New, unverified users in the test face added payment friction before they can participate, while low-value automated account creation becomes somewhat more costly.
The critique highlights that a flat $1 fee is unlikely to be a complete anti-manipulation control when coordinated actors can justify the cost.
Second-order effects
X may need to combine payment gates with stronger behavioral or account-level enforcement rather than treat subscription status as a reliable signal of authenticity.
A fee intended to deter abuse can also filter legitimate low-intent or price-sensitive newcomers, forcing a trade-off between participation and cleaner interactions.
Third-order effects
If platforms increasingly price access as an abuse-control mechanism, trust and safety may become more tightly coupled to paid identity and account privileges rather than open participation.
The durable challenge is the tiered-access model: monetization can fund or gate platform features, but it does not by itself remove the incentives that sustain manipulation.
The trend: Social platforms are testing subscription and access friction as trust-and-safety tools, while discovering that economic deterrence must be matched to adversaries’ expected returns.
Twitter is finally back to A/B testing major changes (that are risky ones). Fun fact on how New Zealand is a very common “test market” for a lot of global companies, esp social media ones. English-speaking market but small enough that if things are messed up, revenue unaffected.
I know there's a lot of negativity on how charging for usage will affect the site, and how this goes very clearly agains free speech (its now paid speech!) The positive thing I see is that - seemingly for the first time since the takeover! - Twitter is running *an experiment*...
INSIGHT: This new move isn't primarily about the $1/year cost. It's intended to be a trivial amount to pay. It's about the credit card + phone number verification requirement. Massively harder to scale networks with 10s of 1000s of bots when *every* single bot will require this
There's been a bunch of “but your service is different” responses so let me clarify: the observation is that my experience reconciles with Twitter's proposal in that attaching a payment impacted abuse. Doesn't matter that the services are different, the premise is the same.
This will be a hugely unpopular thing, however... The premise of attaching a nominal cost to a previously free service in order to combat abuse is exactly what I did with the @haveibeenpwned API keys 4 years ago: https://www.troyhunt.com/... This stopped abuse dead. Not a little.…