@NZParliament, the official account for the New Zealand Parliament on X, has stopped posting to the platform. Previously, it had been used to post routine procedural updates. The Clerk of the House has put a stop to that, according to reports by Stuff, which has staked out a longstanding editorial position against X.
Written in a tone of thinly disguised satisfaction, the story frames the move as part of a broader push for public institutions to abandon the platform altogether. Stuff itself has already left X. The narrative being presented is clear enough: responsible institutions are departing, laggards remain, and Parliament has now joined the morally enlightened.
That narrative is buttressed by comments made by the Clerk to Stuff in explanation of his decision. He said that he had stopped use of the platform because he “could no longer support” it, citing reports that X’s AI chatbot could be used to generate abusive, prurient, morally reprehensible and illegal content (including depictions of minors).1
Stuff quotes the Clerk as saying: “I stopped it because of the news I had seen … I could no longer support the platform … it didn’t feel right for my organisation to use a platform that allowed that to happen.”
The repeated use of the personal pronoun is telling. There is no indication in Stuff’s report that there was any direction from the Speaker or of any justification laid out as a purely neutral operational consideration. It was presented as a matter of individual conscience.
If that is not the whole story, now is the time to let us know.
The problem is that this is not the Clerk’s personal social media account. It is @NZParliament. It exists on behalf of all the House’s MPs and, ultimately, the public. When the explanation is framed in the first person singular, it subtly but clearly positions the decision as an individual moral judgement rather than an institutional determination.
In public administration there is a clear analytical distinction between operational management and political positioning. Operational decisions are justified in practical language. The things that go into operational decisions are issues like resourcing, reach, accessibility, security and consistency with statutory purpose.
An administrator might say a platform is no longer effective, no longer safe for staff, or redundant given the existence of other channels.
But “support” is not operational language. It is moral language. It implies endorsement. It implies alignment. It suggests that mere presence amounts to complicity.
Whether public institutions should remain on X is plainly a contested political question. It intersects with speech, censorship, corporate responsibility, child protection, platform governance and the character of the modern public square. Elected politicians disagree about it. Voters disagree about it. Governments overseas disagree about it.
When the Clerk says Parliament can no longer “support” the platform, he is not simply managing a channel. He is signalling that Parliament’s continued presence would amount to moral endorsement. And that is not a neutral act.
The underlying concerns expressed by the Clerk are, it should be noted, very real. The misuse of AI tools to generate deepfake abuse material is serious and repugnant. But capability of misuse is not a coherent standard for state disengagement.
Microsoft Word can be used to draft extremist manifestos or harassment letters. Photoshop has been used for decades to fabricate explicit images. Email is used for fraud. Mobile phones are used for grooming.
The fact that a tool can be misused does not, by itself, answer the question of whether the state may use it for legitimate, lawful and beneficial purposes.
If capability of abuse were the test, government would have to retreat from much of modern technology. The real debate is about where a vendor’s moral responsibility for what users do with their tools begins and ends, and whether any failure justifies institutional withdrawal. That is not a narrow technical question. It is a political and ethical one that is best made at the direction of the democratically accountable.
And it is not a question for the Clerk of the House to unilaterally settle by reference to his own sense of what he can or cannot “support”.
An added complication is that the retreat by left-leaning media and institutions from X didn’t start with the latest AI scandal at all. It kicked off much earlier, after Elon Musk bought Twitter and made a very public point of trying to turn it into a more free-speech platform, explicitly criticising the old regime for suppressing conservative and populist views. From that moment, X stopped being treated as just another comms channel and became a culture war symbol.
As a result, any decision by a public institution to leave, whatever the stated reason, is inevitably read in political terms: either as a principled stand against a “harmful” platform or as a quiet form of boycott against a space that no longer enforces speech norms preferred by liberal institutions. That’s exactly why Parliament has to be extra careful, because it can’t pretend these decisions land in a neutral vacuum.
I am not suggesting that this is what motivated the Clerk, but I am saying that once X became a proxy battlefield in a wider argument about speech, moderation and political bias, any institutional withdrawal was bound to be interpreted through that lens. In that environment, even a decision taken for sincerely held and defensible reasons will not be received as neutral. It will too easily be accused of being political alignment.
So how should the Clerk have acted on his concerns about reputational risk or moral complicity? The proper course would have looked something like this:
He would first have prepared a formal, neutral assessment setting out the operational risks and statutory considerations, consistent with his duty to manage the Office with integrity, efficiency and concern for the public interest.
He would then have referred that assessment to the Speaker, to whom he is responsible for the management of the Office and who occupies the central governance role in Parliament’s institutional administration.
As the issue carries wider institutional or symbolic implications, the Speaker should have tested it through an appropriate collective forum such as the Business Committee, which the Speaker convenes and chairs and which operates on a cross-party basis.
There is no indication in Stuff’s report that the Speaker was consulted or that any cross-party process was followed. If that did occur, now would be the time to say so plainly.
