DOCUMENT - 1
From:
Paul O"Connell
To:
Paul James
Cc:
[email address]; Michael Mathieson
Subject:
Administrivia
Date:
Thursday, 14 August 2025 9:06:53 am
Hey,
Good to talk yesterday, thanks for taking the time. Following on from that, some matters
of administrivia –
High quality EA support will be required. Is this something DIA can provide? I have
thus far been unsuccessful in convincing MoT to allow my EA to join me (which is
my preference).
If DIA can provide logins and devices (computers and phones) for me, Ben
Hansen (in cc), and Michael Mathieson (in cc), that would be much appreciated.
In terms of software – we’d ideally have AI licences on top of standard installs.
Copilot at a minimum (which might already be in your standard install?), ideally
Gemini and Grok as well (using multiple LLMs and cross checking allows to
control for bias and hallucination).
Access – can Minister Watts’ SPS request that his floor be added to my
Parliamentary Access Card please?
Access to your building (45 Pipitea?) – can I please have access to whatever
bicycle parking and locker / change rooms you have on site please?
What time should we arrive and for whom should we ask upon arrival?
Really looking forward to working with you (all!)
Cheers,
P.
Paul O’Connell
Deputy Chief Executive
Sector Strategy Group
Ministry of Transport
| E: p.o’[email address] | transport.govt.nz
under the Official Information Act 1982
9(2)(a)
Executive Assistant: Anna Northcott | |E: [email address]
MINISTRY OF TRANSPORT
Wellington (Head Office) | Ground Floor, 3 Queens Wharf | PO Box 3175 | Wellington 6011 | NEW
ZEALAND | Tel: +64 4 439 9000 |
Released
Auckland | NZ Government Auckland Policy Office |Level 7, 167B Victoria Street West | PO Box 106238 |
Auckland City | Auckland 1143 | NEW ZEALAND | Tel: +64 4 439 9000 |
Disclaimer: This email is only intended to be read by the named recipient. It may contain information which
is confidential, proprietary or the subject of legal privilege. If you are not the intended recipient you must
delete this email and may not use any information contained in it. Legal privilege is not waived because you
DOCUMENT 1-
have read this email.
Please consider the environment before printing this email.
under the Official Information Act 1982
Released
DOCUMENT - 2
From:
Heather Kirkham [DPMC]
To:
Jonathan Franklin; Hamish Fraser; Dan Suter [DPMC]; Jared Mullen; Simon Rae; ^MBIE: Joe McKay; Jessica
Gorman [DPMC]; Liesle Theron [DPMC]; Mark Talbot [DPMC]; Halia Haddad [DPMC]
Cc:
Janine Smith [DPMC]
Subject:
AI generated non-consensual intimate images
Date:
Wednesday, 14 January 2026 5:50:45 pm
Attachments:
image001.png
You don't often get email from [email address]. Learn why this is important
Classification Removed
Afternoon!
I’ve spoken to a number of you today about the regulatory framework in NZ and
elsewhere on AI generated non-consensual intimate images. This email summarises
what I’ve learned, gives you a chance to correct me where I’ve mis-understood or
misinterpreted, and to chime in on next steps, which are proposed to be to ask relevant
MFAT colleagues from UK, Australia, Canada if they can:
Confirm currently regulatory environment wrt AI generated non-consensual
intimate images
Outline any potential changes to the regulatory environment
Any significant enforcement action underway or being considered wrt AI
generated non-consensual intimate images.
NZ’s current regulatory environment
My understanding is that the current regulatory environment for harmful intimate visual
recordings is governed by:
Films Video and Publications Classification Act
Regulates content not platforms.
Criminalises distribution or publication of objectionable content; including
anything related to exploitation of children or young persons for sexual
purposes regardless of format
DIA working with Chief Censor to confirm that AI generated material subject
to classification, but prima facie would expect it to.
Harmful Digital Communications Act – civil and criminal penalties,
Provides steps for redress for individual affected by the posting of harmful
digital communications – redress largely around taking it down.
Unclear if relates to AI generated images as not explicitly contemplated in
under the Official Information Act 1982
the definition of ‘intimate visual recording’. The opportunity to expressly
include deepfakes at the Justice Select Committee was not progressed in
2021.
Crimes Act (216G – N)
intimate visual recording without consent are criminal – making,
possessing (for purpose of sharing or selling), and publishing / sharing /
selling are prohibited.
Unclear if relates to AI generated images as not explicitly contemplated in
Released the definition of ‘intimate visual recording’
Summary table:
Tool /
Generation
Possession
Distribution /
Platform
publication
Adults
No current
Unclear – intimate visual recordings covered in
DOCUMENT - 2 O
ability to
HDCA and Crimes Act, not clear if extends to
regulate
generated images.
Children
Criminal under FVPCA
Out of Scope
UK regulatory environment
Regulatory environment around AI generated non-consensual intimate images:
Tool /
Generation
Possession
Distribution /
Platform
publication
Adults
Online Safety
Criminal this
?
?
Act 2023
week Data
(Use and
Access) Act
s138
Children
Criminal (?)
UK looking at enforcement action under existing legislation – Online Safety Act 2023
which has duties on social media companies and search services – includes duties on
providers to implement systems and processes to reduce risks their services are used
for illegal activity and to take down illegal content when it appears. Companies can be
fined up to £18 million or 10 percent of their qualifying worldwide revenue, whichever is
greater. Criminal action can be taken against senior managers who fail to ensure
companies follow information requests from Ofcom - the regulator of Online Safety Act.
under the Official Information Act 1982
Ofcom considering action against Grok under the Online Safety Act Ofcom launches
investigation into X over Grok sexualised imagery
Currently creation of image not illegal – but Data (Use and Access) Act 2025 makes the
creation of an intimate image without consent, or asking for the creation of such an
image, an offence. The Act has passed, and this section will be (rapidly) bought into
effect this week. Reference
Australia
Released
Australia: Regulatory environment around AI generated non-consensual intimate
images:
Tool /
Generation
Possession
Distribution /
Platform
publication
DOCUMENT - 2
Adults
Online Safety
TBC
TBC
TBC
Children
Act 2021
Criminal (Acts TBC)
eSafety Commissioner is responsible for the Online Safety Act – regulates platforms,
internet and electronic services based on Basic Online Safety Expectations (BOSE)
heather kirkham
Policy Advisory Group
Department of the Prime Minister and Cabinet
9(2)(a)
E [email address]
The information contained in this email message is for the attention of the intended recipient only and is not necessarily the official view or
communication of the Department of the Prime Minister and Cabinet. If you are not the intended recipient you must not disclose, copy or
distribute this message or the information in it. If you have received this message in error, please destroy the email and notify the sender
immediately.
under the Official Information Act 1982
Released
DOCUMENT - 3
From:
Sophia Kalafatelis
To:
9(2)(a)
; John Michael
Cc:
9(2)(a)
; Jared Mullen
Subject:
RE: Online harm and Grok [UNCLASSIFIED]
Date:
Wednesday, 14 January 2026 4:24:43 pm
Attachments:
image001.png
Thank you
From: 9(2)(a)
Sent: 14 January 2026 16:18
To: John Michael
Cc: 9(2)(a)
; Sophia Kalafatelis ; Jared Mullen
Subject: Re: Online harm and Grok [UNCLASSIFIED]
Hi JP,
No. The general issues around the use of AI in both CSAM offending and detection have obviously
been discussed, but nothing specifically relating to Grok or any one of the thousands of other AI
image generation platforms.
Keep in mind that the 5RD CCEN group is not a policy group though. As a technical research and
development group it focuses on developing and sharing actual investigative tools rather than
higher level policy changes/implementation etc. I'm aware that this is quite different to how other
5RD groups operate.
Out of Scope
Thanks,
9(2)(a)
| Senior Investigator
Digital Child Exploitation Team | Wellington
Department of Internal Affairs Te Tari Taiwhenua
www.dia.govt.nz
From: John Michael <[email address]>
Sent: Wednesday, January 14, 2026 3:54 PM
To: 9(2)(a) under the Official Information Act 1982
Cc: 9(2)(a)
Sophia Kalafatelis
<Sophia [email address]>; Jared Mullen <[email address]>
Subject: FW: Online harm and Grok [UNCLASSIFIED]
Hi 9(2)(a)
Has anything been discussed at 5RD re Grok?
Thanks
JP
Released
From: Sophia Kalafatelis <[email address]>
Sent: Wednesday, January 14, 2026 3:45 PM
To: Jared Mullen <[email address]>; John Michael <[email address]>
Subject: FW: Online harm and Grok [UNCLASSIFIED]
Kia ora Jared and JP,
DOCUMENT - 3
I’ve just received this from our colleagues at MBIE. Policy hasn’t been in contact with the
UK. I wonder whether Digital Safety might have or whether it’s come up through the FCM?
Thank you
Sophia
From: Simon Rae <[email address]>
Sent: 14 January 2026 15:42
To: Sophia Kalafatelis <[email address]>
Subject: Online harm and Grok [UNCLASSIFIED]
Hi Sophia,
I've had an enquiry from the PM's office (Policy Advisory Group) today about Grok. The PM had
picked up a story about the UK consulting Canada and Australia on a potential response to
production of nudes, and wanted to know first whether we had been approached by the UK or
whether we might consider contacting them.
We haven't been contacted, but I was wondering whether (a) you guys had had any contact with
the UK and (b) who the best person in DIA might be for PAG to follow up with.
Simon
Simon Rae
POLICY DIRECTOR
Technology and Innovation, Labour, Science and Enterprise
Ministry of Business, Innovation & Employment/Hīkina Whakatutuki
9(2)(a)
15 Stout Street, PO Box 5762, Wellington, 6145
under the Official Information Act 1982
Released
DOCUMENT - 3
From:
Gillian Ferguson
To:
Nick Law
Subject:
RE: Online harm and Grok [UNCLASSIFIED]
Date:
Friday, 23 January 2026 11:28:25 am
Attachments:
image001.png
Thanks Nick, appreciate that info.
Gillian
From: Nick Law <[email address]>
Sent: Friday, January 23, 2026 11:05 AM
To: Gillian Ferguson <[email address]>
Subject: RE: Online harm and Grok [UNCLASSIFIED]
Hi Gillian
I’ve only spoken to Heather once and it was yesterday afternoon after our 1:1, I wouldn’t
say it was an ongoing discussion with PAG. She called asking for some background on
the current regulatory system, regulators and their roles, regulatory gaps, to test her
understanding of work by other countries and work that has been done here on
regulating platforms. I mostly gave her information that is in the public domain but I did
suggest she accessed the December Cabinet paper on social media regulation and had
a look at the Phase 2 discussion.
We did not specifically discuss the GROK issue. The Grok issue has been flying around
agencies this week. We have not done anything in particular on it and Heather didn’t
ask me about it.
Attached is the email I sent her with links to documents I referred to in the call. I said I
may have something more helpful providing an analysis of how the other countries
systems work.
Nick
under the Official Information Act 1982
Nick Law (he/him) | Policy Manager
Online safety | Racing | National Library and Archives
9(2)(a)
From: Gillian Ferguson <[email address]>
Sent: Friday, 23 January 2026 10:48 am
To: Nick Law <[email address]>
Released
Subject: FW: Online harm and Grok [UNCLASSIFIED]
Hi Nick, can you please share any written material sent to PAG on this, as well as a summary
of the verbal advice so I’m across the issues? I wasn’t aware there had been ongoing policy
contact with PAG and would have appreciated this being raised at our 1:1.
DOCUMENT - 3
Thanks
Gillian
From: Sophia Kalafatelis <[email address]>
Sent: Thursday, January 15, 2026 11:10 AM
To: Simon Rae <[email address]>
Cc: Gillian Ferguson <[email address]>; Nick Law <[email address]>
Subject: RE: Online harm and Grok [UNCLASSIFIED]
Hi Simon,
I’ve checked in with our operational colleagues and neither they nor us in Policy have
had any contact with the UK about this issue. I’ve copied in General Manager Gillian
Ferguson and Nick Law (Policy Manager) who are the best contacts Nick is on leave
returning next week.
Ngā mihi
Sophia
Sophia Kalafatelis (she/her) | Senior Policy Analyst
Online Safety, Racing, National Library & Archives | Policy and Te Tiriti
Department of Internal Affairs | Te Tari Taiwhenua
45 Pipitea Street, Wellington
From: Simon Rae <[email address]>
Sent: 14 January 2026 15:42
under the Official Information Act 1982
To: Sophia Kalafatelis <[email address]>
Subject: Online harm and Grok [UNCLASSIFIED]
Hi Sophia,
I've had an enquiry from the PM's office (Policy Advisory Group) today about Grok. The PM had
picked up a story about the UK consulting Canada and Australia on a potential response to
Released
production of nudes, and wanted to know first whether we had been approached by the UK or
whether we might consider contacting them.
We haven't been contacted, but I was wondering whether (a) you guys had had any contact
with the UK and (b) who the best person in DIA might be for PAG to follow up with.
Simon
Simon Rae
POLICY DIRECTOR
Technology and Innovation, Labour, Science and Enterprise
Ministry of Business, Innovation & Employment/Hīkina Whakatutuki
[email address] | 9(2)(a)
15 Stout Street, PO Box 5762, Wellington, 6145
under the Official Information Act 1982
Released
DOCUMENT - 4
From:
9(2)(a)
To:
Jared Mullen
Cc:
John Michael
Subject:
RE: Media enquiry about public use of X/Twitter
Date:
Thursday, 15 January 2026 7:43:47 am
Attachments:
image002.png
image003.jpg
…and here’s the result… the only mention of DIA being the reporting page!!!
https://www.emilywrites.co.nz/nz-police-dont-know-what-to-do-about-grok-child-sexual-
abuse-images/
9(2)(a)
Senior Investigator
Digital Child Exploitation Team
9(2)(a)
From: Media Internal Affairs
Sent: Wednesday, 14 January 2026 3:34 PM
To: Jared Mullen ; Media Internal Affairs
Cc: 9(2)(a)
; John Michael ; Elizabeth Hare
Subject: RE: Media enquiry about public use of X/Twitter
For sure! Thanks Jared.
Ngā mihi,
Hannah Johns
Senior Communications Advisor
9(2)(a)
From: Jared Mullen <[email address]>
Sent: Wednesday, 14 January 2026 3:34 pm
To: Media Internal Affairs <[email address]>
Cc: 9(2)(a)
John Michael <[email address]>;
Elizabeth Hare <[email address]>
Subject: RE: Media enquiry about public use of X/Twitter
Yes – but please insert the DS reporting weblink at the bottom of the answers.
Jared
under the Official Information Act 1982
From: Med a Internal Affairs <[email address]>
Sent: Wednesday, January 14, 2026 3:27 PM
To: Jared Mullen <[email address]>
Cc: 9(2)(a)
; John Michael <[email address]>;
Elizabeth Hare <[email address]>
Subject: RE: Media enquiry about public use of X/Twitter
@Jared – is this good to go? It’s due 4pm.
Released
Ngā mihi,
Hannah Johns
Senior Communications Advisor
9(2)(a)
From: Elizabeth Hare <[email address]>
DOCUMENT - 4
Sent: Wednesday, 14 January 2026 3:09 pm
To: Media Internal Affairs <[email address]>
Subject: FW: Media enquiry about public use of X/Twitter
Hey Hannah,
See Jared’s response below.
From: Jared Mullen <[email address]>
Sent: Wednesday, 14 January 2026 2:54 pm
To: Elizabeth Hare <[email address]>
Cc: 9(2)(a)
John Michael <[email address]>
Subject: FW: Media enquiry about public use of X/Twitter
Kia ora Elizabeth
Here are our draft responses.
In your organisation's opinion, is it legal to "nudify" pictures of users without consent -
including through tangential methods like "portray her wearing a clingwrap bikini"
If the “nudification” renders images of adult content it may not contravene the Films,
Videos and Publications Classification Act (FVPCA). But these decisions can only be
made on a case by case basis.
There may be offences committed under the Digital Harmful Communications Act or
Crimes Act (in certain circumstances, for harassment).
Is it legal for users to similarly "nudify" or adjust images of children to portray them in
scantily-clad clothing, "clingwrap bikinis" or similar?
Any image which promotes or supports the exploitation of children or young persons, or
both for sexual purposes, or depicts sexual conduct with or by children, will likely be
illegal under the FVPCA.
Does this material constitute child sexual abuse material (CSAM)?
If it meets the above threshold then, yes. CSAM is the term used to describe illegal
material depicting children or young persons.
Given the amount of AI generated CSAM on Grok/X, is it legal for New Zealanders to use
those platforms?
The use of AI tools is not illegal. However, it is illegal for New Zealanders to generate or
spread CSAM on any platform.
under the Official Information Act 1982
If it is legal why?
As with all advanced technology, AI platforms can be (and are most often) used for
legitimate purposes and a blanket ban, would disadvantage the large proportion of
legitimate users.
If it is illegal, why?
Do these images/apps fall under the jurisdiction of the Classification Office, in a similar
manner to films, TV, publications, and video games?
Released
Apps or platforms are not considered to be publications under the FVPCA. However,
images and content generated on apps and platforms are publications for the purposes
of the FVPCA and are subject to decisions by the Classification Office and enforcement
from DIA.
DOCUMENT - 4
Why has X/Twitter/Grok not been subject to restrictions in New Zealand?
New Zealand does not currently have a regulatory framework that covers social media
platforms. The FVPCA, HDCA and Crimes Act cover the actions of New Zealanders
using those platforms.
Major US-AI providers detect and report instances of CSAM production using their
systems as required by US law. Where these referrals identify a New Zealand origin we
will investigate them using our standard practice.
If not, why not?
See above
What should New Zealanders do if they encounter AI-generated CSAM or
nonconsensual "nudified" pornography on those platforms?
Users can use ‘in app’ safety mechanisms to report offending material and to prevent
access to the account responsible for the content they are exposed to.
What should adult members of the New Zealand public do if their image is "nudified" or
made pornographic without consent on Grok/X?
Members of the public could contact Netsafe if a “nudified” image has been sent to
another person or distributed on an online platform as this may constitute a breach of
the Harmful Digital Communications Act.
What should members of the New Zealand public do if an image of their child is used to
generate "nudified" images or similar on Grok/X/any other platform?
The user can report to the platform, and report instances of CSAM to the Department
Digital Safety team at [DS Web Site]
From: Elizabeth Hare <Elizabeth.Hare@dia govt.nz>
Sent: Wednesday, January 14, 2026 11:44 AM
To: Jared Mullen <[email address]>; John Michael <[email address]>;
Tim Houston <[email address]>; 9(2)(a)
Subject: FW: Media enquiry about public use of X/Twitter
Kia ora team,
I hope you are all well.
Please see below pretty lengthy media request we have just received, another one
surrounding Grok.
under the Official Information Act 1982
The way Josh has positioned these questions, I think we can answer them concisely.
Please note the
4pm deadline.
Josh has asked: “If you cannot answer all the questions by the deadline, please answer
which ones you can today and note which questions can be answered later in the week.
If you cannot answer a question, please indicate which ones, and why they cannot be
answered.”
Released
Let me know if this is doable.
Ngā mihi,
Elizabeth Hare (she/her)
Communications Advisor
Department of Internal Affairs | Te Tari Taiwhenua
9(2)(a)
DOCUMENT - 4
dia.govt.nz | Facebook | LinkedIn
From: 9(2)(a)
Sent: Wednesday, 14 January 2026 11:23 am
To: [email address]
Subject: Media enquiry about public use of X/Twitter
You don't often get email from [email address]. Learn why this is important
Hi, my name is 9(2)(a)
. I'm a freelance journalist and my work is published at
a number of outlets including the Spinoff, Webworm, Emily Writes, and my own blog. I
am collaborating with Emily Writes on a piece about public use of and exposure to
potentially illegal content on X/Grok. Multiple articles have been published about this
content: here are two examples.
https://www.404media.co/grok-ai-sexual-abuse-imagery twitter/
https://www.webworm.co/grok/
My questions are:
In your organisation's opinion, is it legal to "nudify" pictures of users without consent -
including through tangential methods like "portray her wearing a clingwrap bikini"
Is it legal for users to similarly "nudify" or adjust images of children to portray them in
scantily-clad clothing, "clingwrap bikinis" or similar?
Does this material constitute child sexual abuse material (CSAM)?
Given the amount of AI generated CSAM on Grok/X, is it legal for New Zealanders to use
those platforms?
If it is legal, why?
If it is illegal, why?
Do these images/apps fall under the jurisdiction of the Classification Office, in a similar
manner to films, TV, publications, and video games?
Why has X/Twitter/Grok not been subject to restrictions in New Zealand?
under the Official Information Act 1982
If not, why not?
What should New Zealanders do if they encounter AI-generated CSAM or
nonconsensual "nudified" pornography on those platforms?
What should adult members of the New Zealand public do if their image is "nudified" or
made pornographic without consent on Grok/X?
What should members of the New Zealand public do if an image of their child is used to
generate "nudified" images or similar on Grok/X/any other platform?
Released
Thanks in advance for your answers to these questions. We anticipate publishing today
at 5pm, so please send your answers by 4pm today. If you cannot answer
all the
questions by the deadline, please answer which ones you can today and note which
questions can be answered later in the week. If you cannot answer a question, please
DOCUMENT - 4
indicate which ones, and why they cannot be answered.
If any questions can be answered on background, my contact phone number is below.
Cheers,
9(2)(a)
under the Official Information Act 1982
Released
DOCUMENT - 5
From:
Jared Mullen
To:
Paul James
Cc:
Karen Hope-Cross
Subject:
RE: Digital Safety Approach to AI-Generated Illegal Content
Date:
Thursday, 15 January 2026 11:55:20 am
Attachments:
image001.png
Thanks Paul.
Eve has this and has circulated it within Minister IA’s Office.
I’ve also provided this – and other background - to Heather Kirkham (DPMC PAG
Advisor for science and technology). Heather was preparing material yesterday for the
PM and she and I had a detailed conversation.
Happy New Year! Lots of interesting issues to look forward to on the Digital Safety front-
no risk of boredom or being idle.
Noho ora mai
Jared Mullen
General Manager Digital Safety and Identity Investigations
New Zealand Department of Internal Affairs
Email: [email address]
9(2)(a)
From: Paul James <[email address]>
Sent: Wednesday, January 14, 2026 6:03 PM
To: Jared Mullen <[email address]>
Subject: FW: Dig tal Safety Approach to AI-Generated Illegal Content
under the Official Information Act 1982
Happy New Year Jared,
Excellent email, if you haven’t already I suggest you proactively share with the Private
Secretaries in Min IA’s office.
Nga mihi
Paul
Released
Paul James
Secretary for Internal Affairs |Tumu Whakarae mō Te Tari Taiwhenua
Government Chief Digital Officer | Secretary for Local Government
Department of Internal Affairs | Te Tari Taiwhenua
DOCUMENT - 5
45 Pipitea Street PO Box 805, Wellington, 6140
dia.govt.nz | Facebook | LinkedIn
From: Fergus Broom <[email address]>
Sent: 14 January 2026 12:30
To: Paul James <[email address]>
Cc: Karen Hope-Cross <[email address]>
Subject: FW: Digital Safety Approach to AI-Generated Illegal Content
Hi Paul
Here’s the email from Jared, as mentioned in my text
Ngā mihi
Fergus
From: Jared Mullen <[email address]>
Sent: Tuesday, 13 January 2026 12:41 pm
To: Rachel Leota <[email address]>; Russell Burnard <[email address]>;
Karen Hope-Cross <Karen.Hope-Cross@dia govt.nz>; Fergus Broom
<[email address]>; Sean O'Neill <Sean.O'[email address]>;
9(2)(a)
Cc: Media Internal Affairs <[email address]>; John Michael <[email address]>; 9(2)(a)
@dia.govt.nz>
Subject: Digital Safety Approach to AI-Generated Illegal Content
Kia ora colleagues
under the Official Information Act 1982
Digital Safety has had a steady stream of inquiries from multiple sources over the past
few days in response to public and media concern about the use of AI to generate
harmful content. I understand that the media has also approached the Minister.
I thought I would set out Digital Safety’s current approach to you all simultaneously, so
that there is reduced risk of confusion.
Released
Context:
The UK technology secretary has backed blocking UK access to X due to alleged
non-compliance with UK online safety laws. In particular, the use of Grok AI in
generating illegal imagery, including child sexual abuse material (CSAM) and
DOCUMENT - 5
deepfake content.
NZ media, including RNZ and Newsroom, have simultaneously queried DIA and
the Minister’s office on the same issue.
We have also received public inquiries on this issue – two demanding that the
Government take action to address illegal content from Grok (X) AI including
banning/blocking the platform.
Mandate:
DIA’s role is narrow and clear under the Films, Videos, and Publications
Classification Act 1993 we are concerned only with illegal content, particularly:
Illegal CSAM
Illegal promotional terrorist content
Both are illegal whether real or AI-generated.
Non-consensual deepfake sexual imagery generally could potentially be captured
under the
Harmful Digital Communications Act or
Crimes Act, but these are
not DIA’s mandate.
Current Situation:
U.S.-based platforms must report CSAM to NCMEC, which refers cases to NZ.
This includes AI generated content.
NZ has received 136 referrals from X/Grok AI from 31 December 2025 to 10
January 2026; these are being triaged and prioritized and will be tackled in the
usual way.
DIA can issue takedown notices, a process routinely applied for terrorist content.
Compliance varies—X is often less cooperative than other platforms.
Key Points:
DIA has
no legal authority to ban X, or Grok AI. This would also disrupt
legitimate AI use.
Filtering social media content via the Child Exploitation Filter is
not the best
approach; most content is accessed in-app and encrypted.
under the Official Information Act 1982
Under current policy settings, our current approach to illegal content on social
media (including AI generated social media content) focuses on
investigation,
takedown notices, and platform engagement, not blanket bans or filtering.
Please let us know if further briefing or talking points are required for media or
ministerial queries.
Released
Jared Mullen
General Manager Digital Safety and Identity Investigations
New Zealand Department of Internal Affairs
DOCUMENT - 5
Email: [email address]
9(2)(a)
under the Official Information Act 1982
Released
DOCUMENT - 6
From:
Karen Hope-Cross
To:
Nick Law; Jared Mullen
Cc:
9(2)(a)
; Gillian Ferguson
Subject:
FW: thank you
Date:
Friday, 23 January 2026 10:42:05 am
Attachments:
image001.png
Morena
Some excellent feedback from our PAG advisor for you, and likely some of your kaimahi.
Well done
Ngā mihi
Karen
From: Heather Kirkham [DPMC]
Sent: Thursday, January 22, 2026 6:15 PM
To: Paul James
Cc: Karen Hope-Cross
Subject: thank you
Classification Removed
Hey Paul,
Happy new year!
9(2)(g)(i)
to understand the
landscape around regulation of tech platforms in the context of the Grok / AI generated
non consensus intimate images that became an issue last week.
They have been excellent - helpful, knowledgeable, thoughtful, concise and most of all
patient with me given my total ignorance and calling out of the blue, particularly Nick
Law and Jared Mullen.
h
heather kirkham
O
Policy Advisory Group
Department of the Prime Minister and Cabinet
9(2)(a)
E [email address]
under the Official Information Act 1982
The information contained in this email message is for the attention of the intended recipient only and is not
necessarily the official view or communication of the Department of the Prime Minister and Cabinet. If you are not
the intended recipient you must not disclose, copy or distribute this message or the information in it. If you have
received this message in error, please destroy the email and notify the sender immediately.
Released
DOCUMENT - 7
From:
Karen Hope-Cross
To:
Paul James
Cc:
Sue Goile; Fergus Broom
Subject:
Classification
OoTS weekly email
Remove
Date:
Friday, 23 January 2026 3:54:14 pm
Attachments:
image001.png
Kia ora Paul
Hope you have had a great break. We have put everything in your tray for your return. This includes
a separate summary of what came up as Acting Secretary.
Please find this week’s update.
Out of Scope
under the Official Information Act 1982
Released
9(2)(g)(i)
AI meeting with Minister Van Velden
You are meeting with Minister Velden on Thursday. The briefing on AI deployment at DIA has gone
across. We’ve also heard through Eve that the Minister has raised questions about Grok AI and the
current regulatory framework, so it’s likely that these broader questions may come up too. We’ve
asked Jared to provide a state of play on this issue (which we have), including any discussions with
other agencies about wider work across government.
Beyond the briefing and note on Grok, is there anything else you think you might need for this
meeting?
Out of Scope
under the Official Information Act 1982
Released
DOCUMENT - 9
From:
Jared Mullen
To:
Fergus Broom; Gillian Ferguson; Murray Davey
Cc:
Nicki Truesdale; Danielle Finnie; Sue Goile; 9(2)(a)
Subject:
RE: Meeting with Minister van Velden - AI use at DIA
Date:
Friday, 23 January 2026 2:33:31 pm
Attachments:
image001.gif
Notes for Paul James - AI-Generated Material - January 2026.docx
Kia ora Fergus and colleagues
As requested – a one-pager on AI is attached.
We haven’t referenced the MBIE policy statements on AI BUT – these were about enabling commercial use and
had an economic growth lens rather than a regulatory or public safety lens.
Noho ora mai
Jared Mullen
General Manager Digital Safety and Identity Investigations
New Zealand Department of Internal Affairs
Email: [email address]
9(2)(a)
From: Jared Mullen
Sent: Friday, January 23, 2026 10:04 AM
To: Fergus Broom ; Gillian Ferguson ; Murray Davey
Cc: Nicki Truesdale ; Danielle Finnie ; Sue Goile ; 9(2)(a)
Subject: Re: Meeting with Minister van Velden - AI use at DIA
Certainly Fergus.
I will also copy in these addressees so that everyone has the info.
Cheers
Jared
Get Outlook for iOS
From: Fergus Broom
Sent: Friday, January 23, 2026 9:16:12 AM
To: Jared Mullen ; Gillian Ferguson ; Murray Davey
Cc: Nicki Truesdale ; Danielle Finnie ; Sue Goile
Subject: FW: Meeting with Minister van Velden - AI use at DIA
Hi both,
See below – sounds like Minister Van Velden is interested in Grok AI and may raise questions at the AI meeting
you and Paul have with her on Thursday next week, Murray.
Jared: I know your team has provided some useful info on Grok over the last week or two and how it is regulated
under current legislation. Are you able to pull this together into a 1 pager for Paul? Wouldn’t need a coversheet
or anything, just over email is fine. Would be good to get an idea of whether there’s anything happening across
government (e.g. with Justice). Could we please get this by
COP Tuesday 27 January?
Gillian: Just looping you in given the Minister is asking about the regulatory framework and where any gaps
might be, and it could come up in the context of social media regulation.
Ngā mihi
under the Official Information Act 1982
Fergus
From: Karen Hope-Cross
Sent: Thursday, 22 January 2026 4:38 pm
To Fergus Broom
Subject: FW: Meeting with Minister van Velden - AI use at DIA
Can you progress this on what Paul may need please.
Ngā mihi
Karen
From: 9(2)(a)
>
Released
Sent: Thursday, January 22, 2026 4:10 PM
To: Murray Davey <[email address]>
Cc: Danielle Finnie <[email address]>; Sue Goile <[email address]>; Ministerial Correspondence
<[email address]>; Rae Kitney <[email address]>; Karen Hope-Cross <[email address]>
Subject: RE: Meeting with Minister van Velden - AI use at DIA
Great, thanks Murray.
FYI ahead of the meeting – the Minister has been quite engaged on the topic of AI across various work
DOCUMENT - 9
programmes and the issue with Grok AI being used to create deepfakes is a matter which is still top of mind.
This is likely to mean the Minister may raise broader queries around the use of AI including current reach of
legislation and what is unregulated.
9(2)(a)
Authorised by Hon Brooke van Velden, Parliament Buildings, Wellington
From: Murray Davey <[email address]>
Sent: Thursday, 22 January 2026 3:48 pm
To: 9(2)(a)
Cc: Danielle Finnie <[email address]>; Sue Goile <[email address]>; Ministerial Correspondence
<[email address]>; Rae Kitney <[email address]>
Subject: RE: Meeting with Minister van Velden - AI use at DIA
Hi 9(2)(a)
Please find attached the briefing for the Minister to support the meeting “AI use at DIA” on the 29 January. Please let me
know if you would like any other information prior to the meeting.
Sue, Dani – Can you please include this for Paul and Myles in their meeting packs
Ministerial Correspondence – For your records. The Ministerial Database number is IA202610832.
Regards
Murray
Murray Davey (he/him) | Chief Digital Officer / GM Enterprise Digital Services
Enterprise Digital Services | Te Ara Matihiko
Digital Services Branch | Te Pūnaha Matihiko
The Department of Internal Affairs | Te Tari Taiwhenua
9(2)(a)
| 45 Pipitea Street | PO Box 1568, Wellington 6140, New Zealand | www.dia.govt.nz
-----Original Appointment-----
From: 9(2)(a)
.go t.nz>
Sent: Tuesday, 2 December 2025 10:58 am
To: 9(2)(a)
; Murray Davey; Myles Ward
Cc: Danielle Finnie
Subject: Meeting with Minister van Velden - AI use at DIA
When: Thursday, 29 January 2026 10:15 am-10:45 am (UTC+12:00) Auckland, Wellington.
Where: EW6.1
You don't often get email from 9(2)(a) [email address]. Learn why this is important
Hi Murray and Myles,
Confirming a time for you to meet with Minister van Velden to discuss AI use at DIA. I have attached Paul’s email
under the Official Information Act 1982
for reference.
As mentioned could you please provide any supporting material for this meeting by 22 January.
Thanks
9(2)(a)
Released
Authorised by Hon Brooke van Velden, Parliament Buildings, Wellington
DOCUMENT - 12
From:
Alison Lewes
To:
Tessa Houghton; Jared Mullen; DS BranchPerformance
Cc:
Frances Parker; Laura Sommer; James Little; Phil Weir
Subject:
RE: Upcoming X (Twitter) meeting with NZ Ambassador to US: request for briefing [SEC=UNCLASSIFIED]
Date:
Wednesday, 28 January 2026 11:59:32 am
Attachments:
RE Upcoming X (Twitter) meeting with NZ Ambassador to US request for briefing SECUNCLASSIFIED .msg
image001.png
Thanks Tessa, that sounds good. @Jared Mullen I will connect you with MFAT (latest
attached) and leave this with your teams.
Cheers
From: Tessa Houghton
Sent: Wednesday, 28 January 2026 11:52 am
Thanks – I think that this meeting is more relevant to MBIE and DIA (Digital Safety) than
GCDO, but there may be some linkages to digital identity re age verification
I’ve cc’d in Jared from our Digital Safety team who has been across conversations about
Grok etc.
Ngā mihi,
Tessa
From: Alison Lewes <[email address]>
Sent: 28 January 2026 11:11
To: DS_BranchPerformance <DS [email address]>; Phil Weir
<[email address]>
Cc: Tessa Houghton <[email address]>; Frances Parker
<[email address]>
Subject: FW: Upcoming X (Twitter) meeting with NZ Ambassador to US: request for briefing
[SEC=UNCLASSIFIED]
Hi there, info request just received for commissioning. I will acknowledge and copy you.
Thanks
From: MURPHY, Emma (ISED) <[email address]>
Sent: Wednesday, 28 January 2026 10:44 am
To: Alison Lewes <[email address]>
Cc: Pedro Ramirez <[email address]>
under the Official Information Act 1982
Subject: FW: Upcoming X (Twitter) meeting with NZ Ambassador to US: request for briefing
[SEC=UNCLASSIFIED]
You don't often get email from [email address]. Learn why this is important
[UNCLASSIFIED]
UNCLASSIFIED
Kia ora Alison,
Released
I’m forwarding the below to you as per Pedro’s out of office.
Best,
Emma
9(2)(a)
UNCLASSIFIED
DOCUMENT - 11
From: MURPHY, Emma (ISED)
Sent: Wednesday, 28 January 2026 10:41
To: '[email address]' <[email address]>; Pedro Ramirez
<[email address]>; Liam Williams <[email address]>
Subject: Upcoming X (Twitter) meeting with NZ Ambassador to US: request for briefing
Kia ora koutou,
Hope you’re all well.
I’m currently covering the Emerging Tech portfolio here in ISED. I’m reaching out as X (fka
Twitter) have requested a meeting with our Ambassador in Washington, and the Post (WSH)
are requesting briefing for this meeting. The meeting date remains TBC, but will likely be
within the next couple of weeks.
We’ve asked WSH for any insights into exactly what X may wish to discuss with the
Ambassador, but we imagine this could cover a wide range of issues from social media age
bans to AI regulation and content issues (e.g. recent issues around sexual sation of children
via X’s AI app).
I hope that MBIE and DIA might be able to contribute to a brief for the Ambassador. If you’re
not the right point of contact for this, grateful if you could please point me in the right
direction.
Thanks in advance for your support.
Ngā mihi nui,
Emma
Emma Murphy (she/her)
Senior Policy Officer, Outer Space & Emerging Technologies
Tupe Whakamaru | International Security & Disarmament Division
Manatū Aorere | New Zealand Ministry of Foreign Affairs & Trade
9(2)(a)
E [email address]
195 Lambton Quay, Private Bag 18901, Wellington 5045, New Zealand
www.mfat.govt.nz | www.safetravel.govt.nz
under the Official Information Act 1982
The information contained in this email message is intended only for the addressee and is not
Released
necessarily the official view or communication of the Ministry. It may be legally privileged. If
you are not the intended recipient you must not use, disclose, copy or distribute this message
or the information in it as this may be unlawful. If you have received this message in error,
please email or telephone the sender immediately ..
DOCUMENT - 13
From:
Jared Mullen
To:
Nick Law; Nikki Hurst; Heather Kirkham [DPMC]; Harris, Lison; ^MBIE: Gina Williamson; Dan Suter [DPMC];
Keys, Fleur; Liesle Theron [DPMC]; Karen McDonald
Cc:
Simon Rae; Mark Talbot [DPMC]; ^MBIE: Joe McKay
Subject:
RE: Platform regulation [UNCLASSIFIED]
Date:
Friday, 30 January 2026 1:39:50 pm
Attachments:
image001.png
image002.png
Kia ora colleagues
Only 1 to 2pm is out for me on Thursday.
Jared Mullen
General Manager Digital Safety and Identity Investigations
New Zealand Department of Internal Affairs
Email: [email address]
9(2)(a)
Out of Scope
under the Official Information Act 1982
Released
DOCUMENT - 13
9(2)(a)
E [email address]
The information contained in this email message is for the attention of the intended recipient only and is not necessarily the official
view or communication of the Department of the Prime Minister and Cabinet. If you are not the intended recipient you must not
disclose, copy or distribute this message or the information in it. If you have received this message in error, please destroy the email
and notify the sender immediately.
From: Heather Kirkham [DPMC] <[email address]>
Sent: Friday, 30 January 2026 10:04 am
To: Harris, Lison <[email address]>; ^MBIE: Gina Williamson
<[email address]>; Jared Mullen <[email address]>; Dan Suter [DPMC]
<[email address]>; Keys, Fleur <[email address]>; Nick Law
<[email address]>; Liesle Theron [DPMC] <[email address]>; Karen
McDonald <[email address]>
Cc: Nikki Hurst <[email address]>; Simon Rae <[email address]>; Mark
Talbot [DPMC] <[email address]>; ^MBIE: Joe McKay <[email address]>
Subject: RE: Platform regulation [UNCLASSIFIED]
Absolutely. I think that would be useful.
From: Harris, Lison <[email address]>
Sent: Friday, 30 January 2026 9:55 am
To: ^MBIE: Gina Williamson <[email address]>; Heather Kirkham [DPMC]
<[email address]>; Jared Mullen <[email address]>; Dan Suter
[DPMC] <[email address]>; Keys, Fleur <[email address]>; Nick Law
<[email address]>; Liesle Theron [DPMC] <[email address]>; Karen
McDonald <[email address]>
Cc: Nikki Hurst <[email address]>; Simon Rae <[email address]>; Mark
Talbot [DPMC] <Mark.Talbot@dpmc govt.nz>; ^MBIE: Joe McKay <[email address]>
Subject: RE: Platform regulation [UNCLASSIFIED]
I agree, it will be useful to see the lay of the land. Would you like us to include someone from
the team which manages the contract with NetSafe?
From: Gina Williamson <[email address]>
Sent: Friday, 30 January 2026 9:28 am
To: Heather Kirkham [DPMC] <[email address]>; Jared Mullen
under the Official Information Act 1982
<[email address]>; Dan Suter [DPMC] <[email address]>; Harris, Lison
<[email address]>; Keys, Fleur <[email address]>; Nick Law
<[email address]>; Liesle Theron [DPMC] <[email address]>; Karen
McDonald <[email address]>
Cc: Nikki Hurst <[email address]>; Simon Rae <[email address]>; Mark
Talbot [DPMC] <[email address]>; Joe McKay <[email address]>
Subject: RE: Platform regulation [UNCLASSIFIED]
Mōrena Heather, kia ora koutou,
Released
Thanks Heather for the suggestion and getting the ball rolling. I’d appreciate the opportunity
to connect with you all to join dots and understand perspectives across agencies.
I’m not able to do 10am Thursday 5 Feb unfortunately, though possibly other times that day /
week and week following. @Karen, might you kindly assist us in finding a time that suits for
this cross-agency kōrero?
Out of Scope
under the Official Information Act 1982
Released
DOCUMENT - 14
From:
Natalia Fareti
To:
Gillian Ferguson; Nick Law
Subject:
Re: AI Issues
Date:
Friday, 30 January 2026 2:44:51 pm
Attachments:
image001.jpg
Noted and yes happy to. We'll chat early next week Nick - back-to-back engagements
today!
Ngā mihi,
~ Natalia
From: Gillian Ferguson <[email address]>
Sent: Friday, January 30, 2026 12:31:31 PM
To: Natalia Fareti <[email address]>; Nick Law <[email address]>
Subject: AI Issues
Kia ora kōrua
Following on from this week’s officials meeting with Minister Stanford where she asked for
advice on how other jurisdictions are approaching the Grok issue, could you please work
together to consider the advice that DIA has already prepared for Minister van Velden on this
area and how this is best re-purposed to meet Minister Stanford’s needs. I’m comfortable
with either a stand-alone info briefing to Minister Stanford or integrating this into the next
project update. Under either approach this is something that I’ll sign out in my role across
both areas please.
Ngā mihi
Gillian
Gillian Ferguson | GM Poli y
Policy and Te Tiriti Branch
Department of Internal Affairs | Te Tari Taiwhenua
9(2)(a)
www.dia.govt.nz
under the Official Information Act 1982
Released
DOCUMENT - 16
Out of Scope
From: Michael Stuart
Sent: 03 February 2026 11:02 AM
To: Pedro Ramirez
Cc: Frances Parker ; Phil Weir ; Besa Chembo ; Andy Cameron
Subject: Re: FOR STRATEGY & EDS REV EW APPROVAL: Comms pack Post interview Paul James -
Public Service AI Work Programme February 2026
Welcome back Pedro.
Yes Phil has provided helpful feedback and comments in the
Reactive messaging section, thanks
Phil. I've updated as recommended.
There's an outstanding review re: ManageMyHealth which Laura will look at.
under the Official Information Act 1982
I've sent it to Murray Davey too for review of the messaging re: internal DIA AI programs, and one
question about Grok specifically, tho of course he should be across the whole memo as the DIA
Ch ef Digital Officer and senior responsible official for AI.
The toolkit was released last week so I've added the following messaging sourced from
Digital.govt.nz to help address the
How is the public sector workforce building AI skills and
digital capability? question.
Released
Additionally, we have just released the Public Service AI Toolkit. It consolidates resources to
support agencies and their staff in planning, implementing and managing AI technologies
safely and responsibly.
The AI Toolkit is a collection of public service-focused standardised resources. The
DOCUMENT - 14
resources are designed to support agencies making investment decisions, as well as
designing, implementing and assuring AI-enabled services.
Currently find in the AI Toolkit are the Use of Artificial Intelligence Policy template and
guidance on records management in the context of AI.
If that's ok by you then I don't think there's any else required from Strategy. Thanks to you, Phil,
Fran and the rest of the team for your invaluable assistance.
Suggested talking points section looks pretty well finalised, and that's the most important bit.
We're on track to submit the memo to OotS by the Monday 10am deadline tho I'll aim to get it
wrapped up on Friday.
I'm in soggy Auckland right now, back tonight. Just a reminder I'm on leave next week but you'll be
in the capable hands of Besa (CCed) for the prep session and the interview itself next week.
Michael Stuart | Senior Communications Advisor
Department of Internal Affairs | Te Tari Taiwhenua
9(2)(a)
45 Pipitea Street | PO Box 805, Wellington 6140, New Zealand
dia.govt.nz | Facebook | LinkedIn
Out of Scope
under the Official Information Act 1982
Released
DOCUMENT - 18
From:
Nick Law
To:
Gillian Ferguson; Sophia Kalafatelis; Mia Alejandro; Natalia Lu
Subject:
RE: For Review - Grok ministerial
Date:
Tuesday, 17 February 2026 1:45:18 pm
Attachments:
image004.png
image005.png
image006.png
image007.png
Thanks Gillian Sophia will take a look.
Nick Law (he/him) | Policy Manager
ational Library and Archives
9(2)(a)
From: Gillian Ferguson <[email address]>
Sent: Tuesday, 17 February 2026 1:12 pm
To: Nick Law <[email address]>; Sophia Kalafatelis <[email address]>; Mia
Alejandro <[email address]>; Natalia Lu <[email address] nz>
Subject: FW: For Review - Grok ministerial
Kia ora tātou, would someone in your team have time to review this response today please? It
doesn’t appear to have come through Policy and there’s content on the Deepfake Bill that
needs some changes (it is presented as a Govt Bill).
Thanks
Gillian
From: Jess Watt <[email address]>
Sent: Tuesday, February 17, 2026 1:05 PM
To: Gillian Ferguson <[email address]>
Subject: Re: For Review - Grok ministerial
Hi Gillian,
under the Official Information Act 1982
Apologies for the short notice, was just advised you’d like to see all grok ministerials.
It’s due to the MO today. It has been through Stacey, just needs to go to Mark for final
sign out.
Ngā mihi
Released
Jess Watt
Senior Advisor, Ministerial and Official Correspondence
Policy & Te Tiriti Branch
Te Tari Taiwhenua Department of Internal Affairs
45 Pipitea Street | PO Box 805, Wellington 6140, New Zealand | www.dia.govt.nz
under the Official Information Act 1982
Released
DOCUMENT - 19
From:
9(2)(a)
To:
Jess Watt; 9(2)(a)
Cc:
Jared Mullen
Subject:
RE: COR3973 RE: GROK APP for DIA
Date:
Thursday, 12 February 2026 9:02:30 am
Attachments:
image001.png
No worries Jess – we will input some standard lines we have used previously into the letter
From: Jess Watt
Sent: Thursday, 12 February 2026 8:52 am
To: 9(2)(a)
Subject: FW: COR3973 RE: GROK APP for DIA
Kia Ora team,
I've received the below ministerial on grok, and am needing help with some content, are you
able to please help?
If you can, please drop in some lines to the doc by
COP Monday?
9(2)(a)
We’ve noticed an increase in these types of correspondence so it may be good if we could get
some standard lines we can use
Ngā mihi
Jess Watt
Senior Advisor, Ministerial and Official Correspondence
Policy & Te Tiriti Branch
Department of Internal Affairs Te Tari Taiwhenua
45 Pipitea Street | PO Box 805, Wellington 6140 New Zealand | www.dia.govt.nz
-----Original Message-----
From: 9(2)(a)
Sent: Friday, 16 January 2026 10:08 AM
To: MPTamaki <[email address]>; Rt Hon Christopher Luxon
<Christopher.Luxon@parliament govt.nz>; Hon David Seymour
<David.Seymour@parliament govt.nz>
Cc: Hon Erica Stanford <[email address]>; Hon Nicola Willis
<[email address]>
Subject: GROK APP
Hello Hon Brooke van Helden (copying on others too)
Please do not wait for your “officials” to look at this gross business. It’s all over the media in Uk
under the Official Information Act 1982
etc. see Newsnight BBC. And listen to Woman’s Hour on BBC Radio 4. Young girls have
committed suicide over being abused by use of AI - shocking.
Needs to be stopped, nipped in bud but everyday we let these dreadful folks get away with their
nasty horrible stuff more young women and even younger are affected. Need I Really say
more??!!
Regards
9(2)(a)
Taupo & Wellington
Released
Sent from my iPhone
DOCUMENT - 20
9(2)(a)
Dear 9(2)(a)
Thank you for your email of 16 January 2026, regarding the use of the Grok app.
As Minister responsible for the Films, Videos, and Publications Classification Act 1993, I am
concerned about all reports of online child sexual exploitation material (CSEM). It is illegal in
New Zealand to generate, possess or share such material including AI generated imagery
using tools like Grok.
Investigators from the Department of Internal Af airs Digital Child Exploitation Team are
dedicated to stopping the spread of CSEM online, identifying and safeguarding victims and
holding perpetrators to account.
Regarding harmful images which do not depict children, you may be aware of MP Laura
McClure’s members bill that proposes to address this issue (the Deepfake Digital Harm and
Exploitation Bill). This Bill seeks to restrict the creation, sharing and threatening to share of
nonconsensual sexually explicit deepfakes. The Bil may be referred to Select Commit ee
this year. I would encourage you to also share your thoughts on this issue through the select
committee submissions process.
Thank you again for writing.
Yours sincerely,
under the Official Information Act 1982
Hon Brooke van Velden
Minister of Internal Affairs
Released
