Yesterday this blog sought to make sense of the decision to drop the Chinese Spying prosecutions on the information then available, and averred that both the positions of the government and the Crown Prosecution Service did not make sense.

Now the government has published three witness statements, the picture now changes. The government’s position now makes sense, and the CPS looks as if it made at least one serious mistake and possibly another.

*

The ‘Joy of Witness Statements’ is a book yet to be written.

*

One always has to be careful with witness statements. They are documents created for a specific purpose: that is to put witness evidence before a court.

They are not essays or expositions to be read generally. They are not literary texts, as such. If there is something in a witness statement not relevant to something which needs to be decided by a court then it usually should not be in a witness statement.

And so when looking at any witness statement the key question is: what was the court being required to decide? You then work backwards from that question to see what is (and what is not) in a witness statement.

Witness statements can be read from beginning to end, like some linear narrative. But that is not the purpose of witness statements. They are complex documents, not linear.

Each paragraph – and each sentence – of a witness statement should related to something which needs to be put before a court.

As such they are documents more akin to computer programs, allowing the court to perform a certain task.

(One common fault of political comentators and literary historians is to use the content of witness statements without regard to what information was needed and not needed to be put before a court.)

*

Yesterday the government of the United Kingdom took the extraordinary step of publishing three witness statements. These are documents which the public rarely get to see. Indeed unless the evidence is repeated in open court, the public would rarely get to know the evidence in such witness statements.

This is especially so in respect of national security matters, where a culture of secrecy means such documents are restricted.

Regardless of any other point, the publication of these statements is a boon for the public understanding of law and should be welcomed.

The government should do it more often.

*

What do these witness statements tell us and not tell us?

First they tell us that the Crown Prosecution Service thought it needed to put evidence before the court.

In particular, as set out in paragraph 3 of the first witness statement dated 22 November 2023, the witness appears to have been asked to provide an “assessment of whether providing information of the type that was alleged to have been provided […] was prejudicial to the safety or interests of the [United Kingdom].”

Here the witness is, of course, echoing what he has been asked to provide.

But.

The statutory wording he is echoing is from the National Security Act 2023.

However, the defendants were not being charged under the 2023 Act but under the Official Secrets Act 1911. The 2023 Act was not in force at the time of the alleged offences.

This first witness statement thereby looks as if the witness was asked the wrong question by the CPS.

The required evidence was not that the material “was prejudicial to the safety or interests of the [United Kingdom]”.

The witness statement had to put evidence in relevant to the 1911 offence – material “which is calculated to be or might be or is intended to be directly or indirectly useful to an enemy”.

So: presumably there was a tactful exchange between the CPS and the government and a new witness statement was required.

Was China an enemy?

You will see that paragraph 4 of the second statement now discusses whether China was a “threat”.

The following paragraph 5 then seeks to substantiate the assertion that China is a “threat” and paragraph 6 then rows that back a little.

This second witness statement was dated 21 February 2024.

The charges are then announced on 22 April 2024.

*

On 16 July 2024 the Court of Appeal hands down its decision on another Official Secrets Act 1911 case, which sets out that one test for the offence is whether the country in question “represents a current threat to the national security of the UK”.

In a letter to MPs, the DPP says of this period:

“In the light of this new judgment, it was considered that further evidence should be obtained. Efforts to obtain that evidence were made over many months, but notwithstanding the fact that further witness statements were provided, none of these stated that at the time of the offence China represented a threat to national security, and by late August 2025 it was realised that this evidence would not be forthcoming. When this became apparent, the case could not proceed. This was a professional assessment made by CPS lawyers experienced in prosecuting national security and espionage cases, applying the Code for Crown Prosecutors.”

*

We now come to the third witness statement dated 4 August 2025 – over a year later from the Court of Appeal decsion.

It can be seen that the CPS had now asked the witness to substantiate that China “represents a current threat to the national security of the UK”.

And so the witness sets out more detail on that point, even adding a sub-heading:

You will see the word “threat” in paragraphs 4 and 7.

Paragraphs 5 and 6 instead use the word “challenge” – but even in paragraph 6 there is mention of “economic power to coerce”.

You will also note that, contrary to news reports, paragraphs 4 and 7 show the witness was not only applying the previous government’s policy, but was also giving his own assessment of the threat of China at the relevant time.

The witness again adds a paragraph rowing back a bit, but not much.

*

So the CPS now not only had paragraph 5 of the second witness statement, it also had paragraphs 4 to 7 of the third witness statement.

The question for the CPS was whether paragraph 5 of the second witness statement and paragraphs 4 to 7 of the third witness statement were enough to show China “represents a current threat to the national security of the UK” at the material time.

You may think that this evidence does, in spades.

But the CPS took the view that not even paragraphs 4 to 7 of the third witness statement were enough to meet the evidential test for there to be a realistic prospect of a conviction.

And so the prosecution was dropped.

But the content of paragraphs 4 to 7 of the third witness statement is hard to square with the DPP’s statement to MPs that “notwithstanding the fact that further witness statements were provided, none of these stated that at the time of the offence China represented a threat to national security”.

*

The CPS decision that the evidence in the witness statements, especially in the third statement, did not meet the evidential test for a prosecution seems perverse.

If this evidence did not meet the test, then what would?

*

The overall impression given by the witness statements is that the CPS position looks muddled. It still makes no sense.

They appear to have used the wrong law when getting the first witness statement, asking the witness to provide evidence in respect of the 2023 offence and not the actually applicable 1911 offence.

Although this error seems to have been corrected by the time of the second witness statement, it raises the question as to whether the CPS applied the correct law when making the charging decision.

If they got the law wrong at that stage, that would explain how somehow the evidential test was “met” at charging stage but not later on.

In any event, the decision to drop the prosecution for evidential reasons makes no sense in view of the third witness statement.

*

As for the government, it looks like the allegation that the government wrongly stuck to the policy of the previous government is not well-founded.

And it looks like the government did what it could to provide the CPS with relevant evidence.

As such, what looked like a nonsensical position yesterday for the government becomes, well, sensical.

Some may say that the witness could have been more robust in his second and third witness statements as to the threat of China.

But there is nothing to indicate that he was undermining or compromising that view. Indeed paragraphs 4 to 7 of the third witness statement provide a plain account of a then-current threat to national security.

There may be evidence which we have not seen about the government’s position, and of any interventions by ministers and officials, but on the basis of the additional information now available from the three witness statements, the fault for the collapsed prosecution lies with the CPS, and not the government.

Usually when one reads up about a confusing-looking case the less confusing it becomes. The is usually because there will be contextual points which make odd looking things seem less odd.

But sometimes the more you find out about a case the more confusing it becomes.

And the dropping of the Chinese spying prosecutions is very confusing indeed.

It simply does not make sense, on the information available.

*

This is not a foreign policy blog and I have no expertise about China as a matter of foreign and security policy. The two defendants in the now-dropped case maintain their innocence, and nothing in this post suggests otherwise.

But this is a blog which often features and explores legal messes – and this is a god-awful legal mess.

Yet legal messes are often instructive – and looking at them can help the public understanding of law.

Seeing how something goes wrong in practice lets us understand how law operates in the real world.

*

When the prosecution was first dropped I noticed the use of a certain phrase.

So I wrote:

The only point this blog will emphasise is that the prosecution was dropped because the “evidential threshold” could not be met.

This is an interesting detail.

*

Why did this detail stand out?

Well, because it is not normal.

By way of background, a prosecution has to meet two “code tests”.

The first is the evidential test – does the available evidence make out the elements of the offence so that there is a realistic chance of conviction?

The second is the public interest test – regardless of the evidential threshold being met, is the prosecution still in the public interest?

If both tests are met then there can be a prosecution.

And both tests are (supposedly) kept under review as the prosecution progresses.

But in practice, there is usually no significant movement on either test between the original charging decision and trial. It is rare for there to be a shift.

*

This prosecution, which had initially met both tests, was dropped because it no longer met the evidential test.

This was strange.

It is rare – especially in a major case – for the evidence base to change so radically from the initial charging decision to the eve of trial so that the evidential test could no longer be met.

This is not the normal course of events.

What had changed?

*

My initial view (which may still have something going for it) was that the evidential test was being used to do the job of the public interest test, but for some reason the public interest had been ruled out.

Sometimes legal messes happen because some area of law or practice is being used (misused, abused) for a purpose for which it is neither fitted nor intended, and what then happens is the unhappy consequence of a misconceived and/or illiberal move.

The Crown Prosecution Service and the government have certainly got themselves in knots in how re-applying the evidential test brought this case to an end. Would it not have been smoother just for the public interest test to be re-applied if geo-political stuff about Chinese threats had changed?

Few would perhaps gainsay an evaluation of the public interest in a national security case like this.

But no: the evidential test was used, and – as we will see – it makes no sense.

*

There is nothing in the news reports and official statements regarding this case to suggest that the direct evidence against the defendants had changed.

The documentary, witness and (presumably) expert evidence against the defendants appears to be the same on the day the case was dropped as on the day they were charged.

Sometimes key direct evidence becomes viewed as inadmissible; or there may be fresh evidence (either way); or the soundness of an expert opinion may be shaken. It happens – rarely – but it happens.

But that does not seem to have happened here.

Nor does there seem to be an exclusion of evidence on the basis of public interest immunity.

The direct evidence base in respect of what the defendants personally did and did not do seems unchanged.

*

On the face of it what changed was not evidence about the defendants but about the “enemy”.

Here we need to look at the relevant statutory offence.

The offence charged was under the Official Secrets Act 1911:

The wording is somewhat out-of-date – “sketch” etc – and the Act has now been replaced with a more modern statute. But this was the offence that in force at the time.

The key phrase for us is “directly or indirectly useful to an enemy”.

*

One may take the view that this reference is not to any particular enemy, but to any enemy. The enemy does not even need to be a nation state.

Imagine, for example, a mercenary spy who takes sensitive information with the view of selling it to the highest bidder, whether that be any enemy state or not (or even back to the government). That information would still be of use to an enemy: we just would not know which one.

*

The view, however, seems to be that there has to be evidence as to the enemy.

That evidence would usually be in the form of a witness statement from an official saying, yes, this is information useful to an enemy.

And if the jury accepts that, and the other elements of the offence are made out, then the jury can convict.

*

But what happened in this case is set out by the Director of Public Prosecutions – the head of the CPS – in a letter to two parliamentary committees:

It is the legal responsibility of the CPS to independently assess the evidence in every case. That responsibility is an ongoing obligation throughout the life of a case. I am satisfied that the decision to charge this case in April 2024 was correct. This was on the basis of where the law stood at that time in relation to the requirements of the Official Secrets Act 1911. Some weeks later, a High Court decision (R v Roussev and others 2024) ruled that “enemy” for the purposes of the 1911 Act includes a country which represents at the time of the offence, a threat to the national security of the UK.

In the light of this new judgment, it was considered that further evidence should be obtained. Efforts to obtain that evidence were made over many months, but notwithstanding the fact that further witness statements were provided, none of these stated that at the time of the offence China represented a threat to national security, and by late August 2025 it was realised that this evidence would not be forthcoming. When this became apparent, the case could not proceed. This was a professional assessment made by CPS lawyers experienced in prosecuting national security and espionage cases, applying the Code for Crown Prosecutors.

*

This account superficially makes sense, and it is written in that nod-along way which can be a feature of such letters.

But with a close look what is said here is problematic.

*

First, note the deft switch from the first person to the passive voice: “I am satisfied that the decision to charge this case in April 2024 was correct. […] it was considered that further evidence should be obtained.”.

Considered by who?

Such a switch in a formal document is usually evasive, and it should put the reader on alert.

*

Second, the DPP refers to “a High Court decision”.

What he appears to mean is not a High Court decision but a decision of the Court of Appeal on appeal from the Crown Court.

Mixing up such courts is something which is fair enough for a lay person but it is something which is strange in a DPP, especially as generally the High Court deals with civil matters.

Perhaps it is an innocent slip, but it also suggests that the DPP may not be that personally familiar with the case and was reliant in the letter on what someone had told him.

Curiously, in Parliament the security minister in a statement on the case also made the very same slip relying on the DPP letter:

What has changed is the CPS’s assessment of the case law. The DPP has explained that in a separate case—the Crown v. Roussev—the High Court ruled on the threshold for evidence needed to prosecute under the antiquated 1911 legislation.

Is nobody actually looking at the cases on which they are seeking to rely?

*

Third, and this is the main point: the Court of Appeal decision widens, not narrows, the meaning of the provision.

The President of the King’s Bench Division said in a joint judgment (broken into smaller paragraphs):

In our judgment, the 1911 Act must be construed without reference to the 2023 Act and in a common sense way. This is not a case of applying any “updated construction”: enemy means the same thing now as it did in the years before the First World War.

We agree with the judge that Russia would be “an enemy” if the jury concluded on the evidence that it was a country with whom the UK might some day be at war. However, neither Phillimore J nor the judge was laying down a comprehensive test.

There is no reason in our view why the term “an enemy” should not include a country which represents a current threat to the national security of the UK.

That formulation may well involve issues of fact and degree which the jury would be well-placed to assess, on evidence. As the judge correctly observed, friendly powers would fall outside this definition.

*

All that the prosecution would need to show with the now-dropped case was that China “was a country with whom the UK might some day be at war”. And it would be for the jury to decide if that were the case.

Note the word “include” by the judge. This means that showing a country “represents a current threat to the national security of the UK” is just one way of showing a country is an enemy for the purpose of the Act.

And as the Court of Appeal says, that “formulation may well involve issues of fact and degree which the jury would be well-placed to assess, on evidence”.

In essence: a wide defintion, and it is for a jury to decide.

*

This is not just my view.

My fellow legal commentator Joshua Rozenberg, with whom I sometimes disagree on international law-related matters, also is unimpressed:

I am not convinced by this reading of the judgment. It seems to me that the Court of Appeal was saying that threatening the UK’s national security was simply one of the ways of establishing that a state could be an enemy. There could be others. If anything, the Court of Appeal was expanding the definition and making it easier to bring charges.

In any event, there is ample evidence that China was a threat to national security at the time. Writing in March 2023, the prime minister Rishi Sunak said that “China poses an epoch-defining challenge to the type of international order we want to see, both in terms of security and values.”

*

And the estimable Professor Mark Elliott states in one of his two excellent posts on this case (here and here):

The effect of this judgment is to make it clear that pre-existing case law that said that ‘enemy’ includes ‘a potential enemy with whom we might some day be at war’ as well as states with which the UK is currently at war was not limiting the category of ‘enemies’ to such states. Thus, the Court of Appeal held in Ivanova and Rossev that the category also includes states that represent a ‘current threat to the national security of the UK’. Contrary to some suggestions that have been made, or at least implied, in the media, this does not make prosecuting for this offence more difficult: by making it clear that the definition of ‘enemy’ is a relatively broad one, it makes establishing that aspect of the offence easier.

*

In other words: if there was sufficient evidence on the “enemy” point for the charges to be brought in the first instance, there should be nothing in the Court of Appeal decision to change the position.

*

Now let us go back to the DPP letter:

In the light of this new judgment, it was considered that further evidence should be obtained. Efforts to obtain that evidence were made over many months, but notwithstanding the fact that further witness statements were provided, none of these stated that at the time of the offence China represented a threat to national security, and by late August 2025 it was realised that this evidence would not be forthcoming. When this became apparent, the case could not proceed. This was a professional assessment made by CPS lawyers experienced in prosecuting national security and espionage cases, applying the Code for Crown Prosecutors.

*

So there are two things here to note: the CPS were requesting new evidence and the government was not giving it – or at least not giving what the CPS wanted.

On the basis of the above, it may have been a mistake for the CPS to seek this new evidence.

But, if the CPS was not mistaken, the culpability seems to switch to the government for not providing the requested evidence.

The key sentence is this:

Efforts to obtain that evidence were made over many months, but notwithstanding the fact that further witness statements were provided, none of these stated that at the time of the offence China represented a threat to national security, and by late August 2025 it was realised that this evidence would not be forthcoming.

*

Now let us go back to the ministerial statement:

The deputy National Security Adviser—a senior official with very extensive experience in matters relating to national security—provided a witness statement in December 2023, under the previous Government. Further witness statements were requested and provided in February and July of this year.

All the evidence provided by the deputy National Security Adviser was based on the law at the time of the offence and the policy position of the Conservative Government at the time of the offence. Every effort was made to provide evidence to support this case within those constraints.

This accords with the DPP letter – further evidence was asked for and witness evidence given.

But it appears that the CPS were insisting that the government witness statement spell out that China was “a threat to national security” and the government refused.

The government then blames “constraints” supposedly set by law “at the time of the offence” and policy “of the Conservative Government at the time of the offence”.

*

This is a really odd form of words. It is not for an official to self-censor their evidence on their personal understanding of the law. That is not their role to second guess what a court may say the applicable law was or was not. Whoever came up with that part of the ministerial statement does not understand witness statements.

Nor was the witness statement likely to be directly about the policy of a previous government. The offence is not concerned about policy positions. And a government is not bound by what a previous government assessed to be a threat.

In essence: the supposed constraints seem unconvincing.

Like the DPP letter, but separately, the ministerial statement makes no sense.

And although one can just about see why the DPP made a mistake, on a misreading of the Court of Appeal decision, it is perhaps impossible to think of a good reason why the government makes its mistake.

*

At his blog, Mark Elliott puts the general position very well:

First, there are publications and statements in the public domain (referred to in my previous post on this topic) indicating that the previous government did consider China to be a threat to national security at the relevant time.

Second, the Prime Minister appeared to suggest that the current government was bound by the previous government’s view as to whether China was a national security threat at the relevant time. But there is nothing in 1911 Act or in relevant case law that justifies such a view.

Third, the Prime Minister appeared to suggest that everything turned on whether China had been ‘designated’ as an enemy or national security threat by the previous government at the relevant time. However, the 1911 Act recognises no such concept of designation, criminal liability turning simply on whether the court is in receipt of evidence to the effect that the country in question was a threat to national security at the relevant time.

*

The more one knows about this case, the more confusing it becomes.

The CPS insists on further evidence it does not require, and the government insists it could not give that evidence, even though it could.

Neither side makes sense, and together they make no sense absolutely.

So either the CPS and the government simultaneously get themselves into an utter muddle on a major espionage prosecution.

And such concurrent cock-ups are conceivable.

Or there is another explanation.