The curious section 3 of the new National Security Act

10th March 2026

*

Hello and welcome to The Empty City blog, the new name of which is explained here.

*

A broad and vague provision may be a cause for concern

*

Last week I was asked to write a piece on the arrests of three individuals in respect of Chinese espionage. One of the individuals was noteworthy, partly because they are married to a member of parliament. It was quite a news event and the New Statesman wanted someone to explain the relevant law.

As is my general approach, I did not mention the individuals or comment on the potential merits of any legal aspect of the case. The individuals are presumed to be innocent until proven guilty, and the recent history of espionage arrests indicate that there is a fair chance the cases go no further. The job of the responsible legal commentator in such circumstances is only to set out the applicable law.

*

I assumed this would be straightforward: an exposition of section 1 of the new National Security Act 2023, which more-or-less replaced the old (and infamous) section 1 of the Official Secrets Act 1911.

But when I looked at the Metropolitan Police press release (always look at original documents when you can, rather than relying on reportage) I noticed something curious.

The arrests were not under section 1 of the new Act.

The arrests were instead under section 3 of the new Act.

Three people arrested under National Security Act Three people have been arrested as part of a Counter Terrorism Policing investigation into suspected National Security Act offences. The men [A-C] were arrested on Wednesday, 4 March on suspicion of assisting a foreign intelligence service, contrary to section 3 of the National Security Act, 2023. The country to which the investigation relates is China.

With emphasis added:

“Three people arrested under National Security Act

“Three people have been arrested as part of a Counter Terrorism Policing investigation into suspected National Security Act offences.

“The men [A-C] were arrested on Wednesday, 4 March on suspicion of assisting a foreign intelligence service, contrary to section 3 of the National Security Act, 2023. The country to which the investigation relates is China.”

*

Section 3 of the new Act is in respect of “Assisting a foreign intelligence service” as opposed to section 1 which is in respect of “Obtaining or disclosing protected information”.

It is a fascinating provision, which you should read in full. In essence it provides for two similarly worded but significantly different offences.

First section 3(1):

“(1) A person commits an offence if the person—

“(a) engages in conduct of any kind, and

“(b) intends that conduct to materially assist a foreign intelligence service in carrying out UK-related activities.”

And then section 3(2) (emphasis added):

“(2) A person commits an offence if the person—

“(a) engages in conduct that is likely to materially assist a foreign intelligence service in carrying out UK-related activities, and

“(b) knows, or having regard to other matters known to them ought reasonably to know, that their conduct is likely to materially assist a foreign intelligence service in carrying out UK-related activities.”

Section 3(3) then provides (emphasis added):

“Conduct that may be likely to materially assist a foreign intelligence service includes providing, or providing access to, information, goods, services or financial benefits (whether directly or indirectly).”

*

The section 3(1) offence is an “intention” offence.

If you look carefully, however, you will notice something which is not there. The person does not actually need to communicate anything to the foreign intelligence service. The person merely has to engage in “conduct of any kind” – even if not communicated.

*

The section 3(2) offence is not an “intention” offence.

Here, the person does not even need to intend to assist foreign intelligence service. Indeed, the foreign intelligence service may not even be aware of it.

The conduct only needs to be “likely” to assist, regardless of intention.

They do not even need to know they are doing it, but “ought reasonably to know”.

And again there does not need to be any communication with the foreign intelligence service – the foreign intelligence service may not even be aware of it.

On the face of it, the section 3(2) offence would cover a hobbyist following foreign affairs or even a journalist, “indirectly” providing access to information on (say) their website or reportage, even if they do not intend to assist a foreign intelligence agency, but in the opinion of the courts “ought” to have known that it would assist a foreign intelligence agency.

*
There are some statutory defences to the section 3(2) offence, listed at section 3(7), but these are strictly limited in scope.

Perhaps there are provisions elsewhere in this complex legislation which mitigate the potential effect of this provision – and if so I will update this page.

Perhaps the the courts will anxiously scrutinise any application of the section 3(2) offence when it impacts upon freedom of expression and the right to share and impart information on matters of topical and public interest.

(Ho ho.)

Perhaps there is nothing to worry about.

*
But.

Section 3(2) is a broad and vague provision.

One can see why the government and the security services want a broad “conduct” approach which would capture any imaginative ploy used by the culpable.

But in doing this there is a risk – as elsewhere with terrorism and national security legislation – that the prosecutorial convenience of the government and the security services is at the expense of clarity and free expression rights.

It will be interesting to see how this section 3(2) offence is used in practice.

**

This post is also cross posted at The Empty City substack, which is run in parallel with this blog.

***

Comments Policy

This blog enjoys a high standard of comments, many of which are better and more interesting than the posts.

Comments are welcome, but they are pre-moderated and comments will not be published if irksome, or if they risk derailing the discussion.

More on the comments policy is here.

Notes on gangster states: why legitimacy matters when the state has a monopoly on coercive and lethal force.

12th January 2026

One priority for a modern state – perhaps the highest priority – is for it to keep its citizens safe from harm.

And within that priority is the need for the state to to keep its citizens safe from harm inflicted by the state itself.

A modern state has – or should have – a monopoly on legitimate coercive and indeed lethal force.

By “legitimate” is meant that the force used will have a lawful basis, will be used in accordance with legal rules, is sanctioned ultimately by someone capable of being publicly accountable, and is capable of review by an independent court.

So what happens when this breaks down?

What happens when the it is the state that is inflicting injury and death on its own citizens – or on people in its care?

And what happens when that coercive and lethal force does not appear to have a legal basis and/or is not in accordance with legal rules and/or is sanctioned by those with no accountability and/or is not capable of independent judicial review?

That is: what happens, for any or all these reasons, the use of coercive and lethal use of force does not seem to have any legitimacy?

What happens when the only justification for the use of coercive and lethal use of force – or excuse or pretext – is that might is right?

Well, among other things, you have a fundamental failure in the functioning of a modern state.

For the key word here is “legitimacy”.

Anybody can (seek to) use coercive and lethal force – and many will get away with it.

And if that is possible, then you have a gangster state, an outlaw state, a pirate state – where anything goes by those who can use coercive and lethal force and be protected from any legal or political consequences.

Some may like the idea of such a state – some may read science fiction or fantasy fiction where there are cities or entire societies where might is right.

Or when may look at various places – now and throughout history – where any sense of a legitimate central order has broken down. And what one will often see are massacres and gangsterism.

That is why there is a quid pro quo – the state gets to have a monopoly on the use of coercive and lethal force, in return for that coercive and lethal force being used legitimately.

That is that the force used will have a lawful basis, will be used in accordance with legal rules, is sanctioned ultimately by someone capable of being publicly accountable, and is capable of review by an independent court.

*

Now looking at the United States, there appears to be the free use of coercive and lethal force by the federal agency ICE (Immigration and Customs Enforcement).

Last week Renee Nicole Good was killed by ICE in circumstances that seem to have been murder: a state execution of an innocent person on the streets for no good reason.

At the start of that day three children had a mother and at the end of that day they did not, just because an ICE agent decided to shoot their mother three times in the face.

I have written about this over at Prospect in an article entitled Death in Minnesota – please click here to read it.

Fatalities happen, police shootings of innocent people happen – but what was especially striking about this killing was the response of the federal government and its supporters.

For want of a better word: they celebrated the killing.

The woman somehow deserved it because she was dubbed “domestic terrorist”.

*

Now turning back to the general issue of legitimacy, we can see that this use of lethal (and not only coercive) force appears from from any of those elements which can (in limited circumstances) render it legitimate.

The extra-judicial execution of Renee Nicole Good seems not to have a legal basis, it was not in accordance with legal rules, was sanctioned (even implicitly) by those who seek to evade accountability, and it may not be capable of review by an independent court.

The federal state is resisting working with the local police.

It looks as if the state is confident it can get away with it.

*

But.

The cost of the state getting away with it will be further diminution of the legitimacy of the state having a monopoly of coercive and legal power.

Yes, as a matter of realpolitik, the state can seemingly do as it wishes.

Yet in the medium term, legitimacy matters.

A society is not easy to govern unless there is at least acceptance of the powers of the police and the military: sheer repression is expensive and often unsustainable.

That is why the wise tyrant hides behind a veneer of legalism and constitutionality.

And that is also why gangster states rarely last very long.

***

Comments Policy

This blog enjoys a high standard of comments, many of which are better and more interesting than the posts.

Comments are welcome, but they are pre-moderated and comments will not be published if irksome, or if they risk derailing the discussion.

More on the comments policy is here.

Miscarriages of Justice: the Oliver Campbell case

21st September 2024

What Elon Musk perhaps gets wrong about civil wars being ‘inevitable’ – It is in the nature of civil wars that they are not often predictable

7th August 2024

On the site that is still known as Twitter, Elon Musk has again tweeted that ‘civil war is inevitable’.

The thing is that civil wars are rarely inevitable – at least not in the short- to medium-term.

This is because civil wars occur, almost by definition, where there is some kind of pre-existing polity which has broken down.

This is what makes them ‘civil wars’ as opposed to any other form of human conflict.

*

A polity often has two key features.

The first feature is a means of regulating disputes – political, legal, social, religious disputes, and so on.

The second feature is a means of enforcing order – usually a form of legitimised coercive power.

It is only when a polity fails to resolve a dispute and then does not enforce order that there risks being a civil war.

The presence of perceived contradictions within a given society is not sufficient: a ruler or ruling class can be quite adept at keeping power despite significant domestic discontent.

Some polities – from tyrannies to loose confederations – can exist when with stark differences between those who are governed.

*

In the longer term, there may be a case for a ‘civil war’ being inevitable.

Most political systems will break down eventually, after a century or two.

In what is now Great Britain and Ireland there have been various civil conflicts on and off for hundreds of years.

If Musk waits long enough then there may again be a civil war one day.

But one suspects that is not what he means.

*

What is often meant by those who say civil conflict is inevitable is that they normatively think that civil conflict should be inevitable.

That they believe there should be a civil conflict on socio-economic or ethnic or religious or some other lines within a given polity.

But, to adapt Karl Marx, ‘the history of all hitherto existing society is the history of those in power often staying in power to the frustration of those who want to have a revolution’.

*

Civil wars are rarely inevitable.

And civil wars deliberately brought about and signalled in advance are rarer still.

If the polity fails to resolve the underlying disputes in a given society – including by the means of effecting fundamental constitutional or socio-economic changes or by granting autonomy or independence to a certain part of the polity – then there is the additional hurdle of the state being normally in a strong starting position to enforce and maintain order.

To say that one thinks normatively there should be a civil war is not the same as saying positively that there will be one.

Yes, every polity is capable of collapsing, and a civil war is thereby always a possibility.

But they are often not predictable when they do happen.

And they then only seem inevitable in hindsight.

**

Comments Policy

This blog enjoys a high standard of comments, many of which are better and more interesting than the posts.

Comments are welcome, but they are pre-moderated and comments will not be published if irksome, or if they risk derailing the discussion.

More on the comments policy is here.

How the Met police may be erring in its political insider betting investigation – and why we should be wary of extending “misconduct of public office” to parliamentary matters, even in nod-along cases

28 June 2024

Why the United Kingdom government cannot leave the ECHR without either breaching or re-negotiating the Good Friday Agreement

1st July 2023

*

The overlooked obstacle to the United Kingdom withdrawing from the ECHR

*

From time to time the demand comes from a government minister, or from one of their political and media supporters, for the United Kingdom to leave the European Convention of Human Rights.

This short blogpost sets out the most obvious obstacle for the government in doing this.

The obstacle – if that is the correct word – is the Good Friday Agreement.

*

That thirty-six page document – which is not as read as widely as it should be – contains a number of express provisions in respect of the ECHR:

“The British Government will complete incorporation into Northern Ireland law of the European Convention on Human Rights (ECHR), with direct access to the courts, and remedies for breach of the Convention, including power for the courts to overrule Assembly legislation on grounds of inconsistency.

[…]

“There will be safeguards to ensure that all sections of the community can participate and work together successfully in the operation of these institutions and that all sections of the community are protected, including:  […]

“(b) the European Convention on Human Rights (ECHR) and any Bill of Rights for Northern Ireland supplementing it, which neither the Assembly nor public bodies can infringe, together with a Human Rights Commission

[…]

“The Assembly will have authority to pass primary legislation for Northern Ireland in devolved areas, subject to: (a) the ECHR […]”

And so on.

*

The ECHR is not just mentioned in passing in a recital.

Instead the ECHR is integral to the Good Friday Agreement.

Rights under the ECHR that can be relied upon in Northern Ireland are a fundamental part of the agreement.

It was important to Ireland – and to the nationalist community – that there were rights beyond the reach of Westminster and Whitehall (and Stormont) that could be enforced directly against the state of the United Kingdom, including against the police and security services.

*

When this obstacle is pointed out, sometimes the response is “Aha! Why not just have the ECHR applicable in Northern Ireland?”

Of course, there is nothing in the Good Friday Agreement which expressly requires rights under the ECHR to be directly enforceable elsewhere in the United Kingdom.

But.

Article 1 of the ECHR provides:

It may thereby not be open to the United Kingdom to be a party to the ECHR and pick-and-choose who within its jurisdiction can have the benefit of the rights.

This would be in addition to the political issues about having a further legal “border down the Irish Sea”, which presumably would not be welcome to unionists.

*

Perhaps the government of the United Kingdom could seek to renegotiate the Good Friday Agreement?

This would mean Ireland agreeing that those – especially nationalists – in Northern Ireland should have their existing legal rights against the United Kingdom state removed.

It would also mean Ireland agreeing that it would not be able to take the United Kingdom to court in Strasbourg.

And it would also mean – in practice – the United States and the nationalist community agreeing that legal rights and protections are removed.

This is not at all realistic.

*

And the difficulty cannot be resolved by simply copying and pasting the Convention rights into a domestic statute for Northern Ireland.

For unless the rights are as constructed and interpreted by the Strasbourg court, and unless a disappointed party can petition the Strasbourg court directly, they are not “convention rights” – even if identically worded.

(This is partly why even Dominic Raab’s “Bill of Rights” that was to repeal the Human Rights Act had the convention rights in a schedule and a duty on public authorities to comply with those rights.)

*

Part of the difficulty of Brexit was because some did not know or did not care about the particular situation of Northern Ireland. Some also pretended it was not an issue, but as we now know it needed special care and attention – and it still has not been fully resolved.

Similarly those who believe just leaving the ECHR would be easy may again be overlooking the Irish and Northern Irish dimensions.

And unless the Good Friday Agreement is re-negotiated, the United Kingdom leaving the ECHR would place the United Kingdom in breach in Good Friday Agreement.

Well, at least as long as Northern Ireland remains part of the United Kingdom.

And that would be another story.

*

This post is partly drawn from this earlier blogpost.

***

Comments Policy

This blog enjoys a high standard of comments, many of which are better and more interesting than the posts.

Comments are welcome, but they are pre-moderated and comments will not be published if irksome, or if they risk derailing the discussion.

More on the comments policy is here.

The commercialisation of private prosecutions

19th May 2023

In the Financial Times magazine this weekend – and on their website (though behind a paywall) – is a fascinating and detailed article on the commercialisation of private prosecutions – especially in respect of shoplifting and online counterfeiting.

*

By way of background: usually one way of explaining the difference between criminal law and civil law is that in the former a person is prosecuted by the state, while in the latter a person is sued by another person.

But with private prosecutions, a person can bring criminal prosecutions against another person.

It is an example of the private enforcement of public power.

*

Of course, the hope (if not expectation) is that any abuse of these prosecutions would be dealt with by an impartial and independent court looking out for the public interest.

But such prosecutions are outside of the processes the police have of dealing with incidents, and also outside of the processes of the Crown Prosecution Service have in determining whether a prosecution should be brought.

Yes, it is possible for the Crown Prosecution Service to step in and terminate a private prosecution, but that is exceptional.

So what we have are defendants – whose cases would have been dealt with differently had the police or the Crown Prosecution Service – facing harsher sanctions at the criminal courts.

And this is done as a business, as the Financial Times spells out, for those bringing these prosecutions only get paid if they can apply for public funds at the end of a successful prosecution.

It seems the various shops and businesses which are affected by the criminality in question do not contribute to the costs of the prosecution.

*

The article points to both a justice gap and to a failure to properly fill that gap.

Many of the shops and businesses nod-along with the private prosecutions because they have no confidence in the police and the Crown Prosecution Service, who in turn are not properly resourced.

And as several of those caught up in the private prosecutions have drugs problems, it can even be contended that some of the prosecutions make no real overall difference to the crime levels, just diverting crime elsewhere from the protected shops and businesses.

*

The Financial Times piece is an interesting sideways snapshot of the criminal justice system.

And if you cannot afford to buy the Financial Times tomorrow, and so decide to read it inside the newsagents instead, please do remember not to walk out without paying for the newspaper.

The article will tell you why.

***

Comments Policy

This blog enjoys a high standard of comments, many of which are better and more interesting than the posts.

Comments are welcome, but they are pre-moderated and comments will not be published if irksome, or if they risk derailing the discussion.

More on the comments policy is here.

Today’s Metropolitan Police apology shows they are still failing over Daniel Morgan

10th May 2023

The Metropolitan Police fail and fail again in respect of the murder of Daniel Morgan.

Some of these failures were in the five botched police investigations and the often accompanying corruption.

But some of these failings are in how the Met has dealt with the disclosure of materials to the Daniel Morgan inquiry, which reported in 2021.

In the report, the panel said (emphasis added):

There was not insignificant obstruction to the Panel’s work. At times the contact between the Panel and the Metropolitan Police resembled police contact with litigants rather than with a body established by the Home Secretary to enquire into the case, and to which the Metropolitan Police had promised to make ‘exceptional and full disclosure’.

“The Panel concludes that, despite the express commitment by the Metropolitan Police in the Terms of Reference to support the Panel’s work, the Metropolitan Police did not approach the Panel’s scrutiny with candour, in an open, honest and transparent way, making exceptional and full disclosure of relevant documents. The way in which material was disclosed or withheld had the effect of making the Panel’s work more difficult.”

And so, to the surprise of absolutely nobody, the Met has now admitted to substantial material non-disclosure:

If you are gullible enough to accept the Met’s explanation for this delayed disclosure at face value, do note that there has been no reason given for why materials found in January are now only being acknowledged in May.

(Also note the deft and vague “number of years” – the appointment of the panel was announced in 2013.)

There can be no good reason for the non-disclosure of these materials and for the delay in admitting they exist.

This is simply a continuation of the evasive and obstructive non-disclosure practices of the Met throughout this whole matter.

The Met did not like – and do not like – the concentrated scrutiny that comes with an inquiry such as that conducted by the Daniel Morgan panel.

The Met would much prefer to deal with the short attention spans of time-poor and story-hungry media.

The problems identified by the Daniel Morgan panel, which I set out in this video, are still present in the Met.

Our thoughts should be with Alastair Morgan and the Morgan family at this latest let-down.

***

Comments Policy

This blog enjoys a high standard of comments, many of which are better and more interesting than the posts.

Comments are welcome, but they are pre-moderated and comments will not be published if irksome, or if they risk derailing the discussion.

More on the comments policy is here.

How Prince Harry’s legal case shows how the phone hacking story has returned to the start of a circle

26th April 2023

The news about the royals and hacking, well summarised and analysed by Joshua Rozenberg at his Substack, brings us back to the start of a circle.

For the phone hacking story only came about because of the royals.

The story came about because the Fleet Street press of the time – with their well-connected links with the Metropolitan police and the private investigation mini-industry, and unchecked by fearful politicians – sought access to information from the voicemails of the royal household.

Because the royal household became involved, the matter was passed to different police officers at the Metropolitan Police, who then raided and took compelling evidence from private investigators.

And in Scotland Yard that evidence was stored, and it became relevant to civil claims some years later, and then suddenly the scope and extent of tabloid phone hacking became apparent.

But without the royal household connection, the crucial evidence would not ever have been seized and stored, and without that evidence being available for later litigation, the hacking story may never have emerged.

What happened shows the practical importance of the monarchy to our politics, regardless of constitutional theory and conventional wisdom.

It seems only the monarchy has any autonomous power when the police and the media and the politicians collude.

*

Such crude phone hacking now seems from another age – technologically, culturally, politically, legally.

After the current crop of cases it may well be that the phone hacking litigation comes to an end.

Prince Harry’s various cases will then perhaps be the other bookend to that provided by the original hacking of the royal household telephones.

But as the parties attend hearings at the Royal Courts of Justice in the Strand, the sophisticated surveillance and data retention by the state and technology companies continues at an unimaginable scale, again unchecked by either politicians or the media.

The phone hacking of a media generation ago seems like a garden shed affair compared with a huge urban conurbation of the exercise of “investigatory powers”.

Any abuses and misuses (or even uses) of the current technology will, in turn, probably never come to light so as to horrify.

Unless, of course, the abuses and misuses (and uses) affect the royal household.

And only then, maybe, will we ever get to hear about it.

***

Comments Policy

This blog enjoys a high standard of comments, many of which are better and more interesting than the posts.

Comments are welcome, but they are pre-moderated and comments will not be published if irksome, or if they risk derailing the discussion.

More on the comments policy is here.

The police officers who want to be armed are perhaps the ones who should not be police officers

20th January 2023

There is an old adage: those who want to have power are the ones who should be disqualified from having power.

Similarly, those police officers who want to be armed should be the ones who perhaps should not even be police officers.

This thought is prompted by the examples first of Wayne Couzens and now David Carrick, both of whom were keen to have the status of being able to have a gun.

Neither Couzens nor Carrick, as far as we know, misused a firearm.

But both seemed desperate to have the status of being able to have a gun and perhaps to boast about it, to themselves and others.

And that was a danger sign.

Of course, there is a necessity to have armed police: that is an unfortunate feature of the modern age.

A civilian and entirely unarmed police force belong to a golden age –  an age which probably never existed.

But.

The question is not whether some police are armed, but about how armed police are selected.

And it would seem those who put themselves forward, so that they would not only have the legal right to inflict coercive force on others, but also be able to inflict lethal force, are the constables who should perhaps be thrown out of the police altogether.

Only perhaps the police officers who are chosen by others – their superiors and peers – to have guns should be the constables who are reluctant to be armed.

And if this approach is adopted then the ‘vetting’ process would be a lot easier:

“Do you want to have a gun? You do? You’re dismissed.”

**

Comments Policy

This blog enjoys a high standard of comments, many of which are better and more interesting than the posts.

Comments are welcome, but they are pre-moderated and comments will not be published if irksome.