How not to leave X
New Zealand Parliament leaving X last week provides an excellent case study in how not to go about it.
Instead of a measured operational decision, they’ve ended up in the media, in the sights of politicians, and on the receiving end of some pretty awful abuse on the platform itself.
Leaving a social platform is never just a tactical move for public sector organisations. It carries both perception risk and political risk. And if you don’t control the reasoning behind it, someone else will.
In Parliament’s case, it feels extraordinarily naïve not to have anticipated the backlash.
So what happened?
On Friday, the @NZParliament account posted that it would “no longer be posting updates on X”, directing followers instead to its other platforms or parliament.nz. No reasons were given.
That absence immediately stoked questions - some genuine, many not. And within hours the replies were filled with accusations of censorship, political bias, wokeness, arrogance, and “stacking the deck” ahead of November’s election.
MPs quickly became involved. Winston Peters labelled the decision “moral virtue signalling”. Chris Bishop suggested the Business Committee would want to discuss it.
The abuse followed just as quickly - vile, aggressive, deeply personal, and in some cases threatening. Deepfake videos of the Clerk of the House have already circulated.
None of this was unpredictable.
Explaining the decision later, Clerk of the House David Wilson told Stuff he could no longer support the platform following concerns about X’s AI chatbot Grok being used to generate deepfake nudes and child exploitation material.
“It didn’t feel right for my organisation to use a platform that allowed that to happen,” he said.
That may be a sincerely held view. But once you frame a platform exit as a moral stand, you inevitably invite political interpretation - particularly in a hyper-polarised environment.
And we’ve been here before.
When the Ministry of Health announced last year it was “reducing posting on X” while also launching on Bluesky, it faced immediate criticism, including from the Minister of Health. An OIA later revealed that officials had anticipated the backlash for months and prepared reactive lines. Internally, concerns centred on staff safety, declining effectiveness, and moderation workload - far more operational than ideological.
It’s also important to put X in perspective within the New Zealand public sector.
A decade ago, it was a common and often effective channel for government departments and councils. Even as recently as 2023, it was still used (albeit often half-heartedly) by 23 departments, 24 Crown agencies, and 46 of 78 councils.
Today those numbers sit at 7 departments, 6 agencies, and zero councils.
That decline predates recent deepfake controversies.
Instead, for most public sector organisations, X stopped delivering meaningful return years ago. Engagement dropped, reach declined, and moderation demands increased. Time and resource (precious commodities in small teams) became harder to justify.
The most relevant question for public organisations in deciding whether they use any platform is usually a simple one: is it strategically useful, and is the return on effort appropriate, given the limited time and resource available?
The Free Speech Union has cited X’s New Zealand user numbers at 933,000. DataReportal’s recent figures - based on potential ad reach - put the number at 678,000 and quickly falling. InternetNZ research indicates only around 6% of New Zealanders use X daily - well down the list of platforms.
Historic follower counts are also not the same as active audiences. Saying Parliament has 35,000 followers (as has been pointed out) may be technically accurate but practically meaningless when only a small fraction are active and engaged.
Leaving a platform because your audience has moved on, or because engagement no longer justifies the effort, can be defensible.
Leaving a platform on the basis of a broad ethical argument, without clearly explaining the full operational context upfront, is much harder to defend - particularly when that argument can be framed as political.
In high-scrutiny environments, decisions about platforms are rarely judged on their internal logic alone - they are judged on timing, framing, and perceived motive.
That doesn’t mean the concerns about X are unfounded. The platform has serious issues, and the tone and substance of replies over the weekend only reinforces that reality.
The challenge for public institutions is that all major platforms carry reputational and ethical risk in different ways. Once decisions are framed as moral judgements about one platform, the obvious question becomes: where do we draw the line, and how do we apply that standard consistently?
Parliament hadn’t meaningfully used X for a long time. Most posts were perfunctory updates with extremely modest reach. If they had reduced posting to once a month, or sunset the account, it is unlikely anyone would have noticed.
Instead, a brief announcement without context created a vacuum. And in online spaces, particularly politically charged ones, vacuums rarely remain empty for long.