R v Yu [2025] EWCA Crim 1066: Foundation for Cross‑Examination on Third‑Party Digital Messages and the Prosecution’s Duty to Pursue Defence‑Flagged Inquiries

Introduction

This commentary examines the England and Wales Court of Appeal (Criminal Division) decision in Yu, R. v [2025] EWCA Crim 1066 (10 July 2025). The case arose on an application for an extension of time (129 days) and for leave to appeal against conviction following a jury verdict of two counts of rape at the Bristol Crown Court.

Two referred grounds were argued: first, whether the trial judge erred in preventing cross-examination of the complainant (A) concerning a WeChat message from an account calling itself “Strawberry” demanding money to halt proceedings; second, whether the judge erred by failing to remind the jury that a witness (Y) said that parts of her police statement concerning the dates of A’s complaints were wrong. A third ground was not renewed and played no further part.

While dismissing the two referred grounds as they stood, the Court of Appeal drew sharp attention to a missed investigative step: despite defence requests, the prosecution had not put to A the straightforward attribution questions—whether she operated the “Strawberry” account or was directly or indirectly involved in sending the money-demand message. The Court directed that a new ground of appeal be drafted to address this omission and set a timetable for the Crown’s response, foreshadowing the possible consequences for the safety of the conviction depending on A’s answers.

Summary of the Judgment

• The trial judge’s refusal to allow cross-examination of A about the “Strawberry” WeChat message was upheld. At trial there was “nothing to suggest that A had anything to do with it,” and permitting questioning would have been a “fishing expedition.”
• The ground that the judge failed to remind the jury about Y’s correction of date errors in her police statement was rejected. The jury had not seen the witness statement; the judge’s decision on what to comment on in summing-up was within his discretion, and there was no basis to revisit the matter.
• However, the Court highlighted a serious procedural lapse: despite proper defence requests, the prosecution had not asked A the attribution questions about “Strawberry.” The Court indicated:
  • These questions were cross-examination “as to credit,” meaning the defence would be bound by A’s answers.
  • Had A admitted involvement, cross-examination on the message contents would have been properly open and, depending on the facts, non-permission might have rendered the conviction unsafe.
  • Given the omission, the Court directed the defence to file a new ground focusing on the failure to conduct that inquiry, with the prosecution to respond within 21 days.
  • If A denies involvement, the new ground is likely to fail; if A admits involvement, the respondent may struggle to resist the appeal.

Factual and Procedural Background

The applicant and A are Chinese nationals who met in late 2022. Following social interactions involving a WeChat group, gifts, and a lunch in Bath, sexual activity occurred at the applicant’s home in the early hours of 28 November 2022. It was common ground that oral and vaginal sex took place and that the applicant left visible marks (including a “love bite” known colloquially in Chinese as a “strawberry”). A’s account was that she did not consent; the applicant’s case was that all sexual activity was consensual.

A reported that the applicant threatened to distribute a video of her; the applicant accepted making such a threat but denied rape. A disclosed to a friend (W) and then reported the matter to the police on 29 November 2022. Police later found that two four-minute videos had been deleted from the applicant’s phone and could not be recovered. The applicant was convicted and sentenced to 9 years’ imprisonment on each count, concurrent.

Leading up to a rescheduled April 2023 trial date, a WeChat contact request to the applicant’s account (managed by a friend) was received from “Strawberry,” with messages amounting to a demand of 20,000 RMB to “solve the situation” and “not show in court.” At trial, the defence sought to cross-examine A on that messaging, alleging an attempt to extort money to withdraw the complaint. The judge refused, finding no evidence linking A to the Strawberry account and deeming the proposed questioning a fishing expedition. A and Y gave evidence; W’s statement was read as she had returned to China, a point not pursued on appeal.

Analysis

Precedents and Authorities Cited

No prior case law was cited in the judgment. Two legal touchpoints are expressly noted:

• Sexual Offences (Amendment) Act 1992: The complainant benefited from lifelong anonymity. The court took care to refer to her as “A” and to other witnesses by initials to avoid inadvertent identification.
• Common law principles on cross-examination “as to credit”: The Court emphasised that the proposed questions about “Strawberry” were questions as to credit. In such circumstances, the defence would be bound by A’s answers.

Beyond these, the judgment proceeds on established trial management and evidential principles: (i) a foundational evidential link is required before a party may cross-examine a witness about third-party communications; (ii) courts will prevent “fishing expeditions”; and (iii) summings-up are a matter for the trial judge’s discretion, subject to ensuring the jury is properly directed on the live issues.

Legal Reasoning

1) Foundation before cross-examining on third-party digital messages

The first ground failed because, at the time of trial, there was no evidence linking A to the “Strawberry” WeChat account or message. Without a basic attribution link, questions suggesting that A sought money to drop the case would be speculative and unfair. The “fishing expedition” label here is critical: cross-examination cannot be a vehicle for rummaging in search of a link that has not been established.

However, the Court drew an equally important counterpoint. Defence requests had been made that the prosecution ask A whether she was the operator of the “Strawberry” account or otherwise involved in the message. Those straightforward attribution questions were never asked. The Court recognised that, had A accepted attribution, the evidential landscape would have changed: the defence would then have had a proper foundation to cross-examine on message content. That could have materially affected the trial—so much so that, on that hypothetical, excluding cross-examination might have rendered the conviction unsafe.

Crucially, the Court characterised the attribution questions as cross-examination “as to credit,” with the consequence that the defence would be bound by A’s answers. This has two implications:

• If A denied involvement with “Strawberry,” the line would probably end there; the defence could not call independent evidence simply to contradict her on a collateral credit point.
• If A admitted involvement, the defence could legitimately explore the messaging further as a live credibility issue for the jury.

2) Trial judge’s handling of Y’s statement and dates

On the second ground, the Court noted that the jury had not seen Y’s police statement. Y said in evidence that the statement contained mistakes as to dates, and she identified those mistakes. The court held there was no basis for the trial judge to refer in summing-up to material in a statement not placed before the jury. Moreover, what a judge chooses to comment on in the summing-up is quintessentially a matter of judicial discretion, provided the live issues are fairly and accurately presented. There was no unfairness here.

3) The prosecution’s omission and appellate case management

The judgment’s most notable development is the Court’s recognition of an outstanding investigative question: the Crown did not ask A the attribution questions despite proper defence requests. The Court set a practical remedial course at the appellate stage:

• The defence must draft a new ground focused on the prosecution’s failure to conduct the requested inquiry about “Strawberry.”
• The Crown has 21 days to respond by Respondent’s Notice.
• The Court indicated the likely outcomes: if A denies involvement, “no viable ground” will likely remain; if A admits involvement, “it may be difficult for the Respondent to resist the appeal.”

This approach balances two concerns: guarding against speculative cross-examination at trial while ensuring that reasonable, targeted defence requests for straightforward attribution inquiries are pursued—especially where an affirmative answer could affect trial fairness and the safety of a conviction.

Impact and Significance

For trial practice

• Defence practitioners should formulate applications to cross-examine on third-party digital communications in two stages:
  • First, seek permission to ask threshold attribution questions that establish a link between the witness and the communication.
  • Only if attribution is admitted (or adequately established) should the defence press to cross-examine on the content and its implications.
• Prosecutors should respond promptly to targeted defence requests to put simple attribution questions to witnesses where a plausible link is raised. Not doing so may later imperil the safety of a conviction if the answer would have opened a legitimate line of cross-examination.
• Trial judges remain entitled to shut down “fishing expeditions.” But this decision encourages judges and parties to distinguish between an impermissible speculative foray and a permissible, narrow threshold inquiry that may establish the necessary foundation for further questioning.

For digital evidence in sexual offence trials

• Digital messaging via platforms like WeChat can implicate important credit issues, including alleged attempts to induce withdrawal of complaints. Attribution is often the determinative hurdle. This case emphasises that attribution needs to be put squarely to the witness where feasible.
• The Court’s intervention underscores that relatively low-cost inquiries (a direct question to the complainant) must not be overlooked when they bear on fairness and may alter the evidential picture.

For appellate case management

• The Court’s directive to create a new ground and set a timetable for the respondent’s notice illustrates a pragmatic, fairness-oriented use of appellate powers to address a discrete, potentially outcome-changing omission without pre-judging the merits.
• The Court signalled clear probable outcomes to guide the parties: a denial likely ends the issue; an admission probably leads to concession or a retrial. This provides clarity and reduces unnecessary procedural churn.

Complex Concepts Simplified

• Cross-examination “as to credit”: Questions aimed at a witness’s honesty or reliability (rather than the core facts) are generally considered collateral. If the witness denies a collateral allegation, the cross-examining party is usually “bound by the answer” and cannot call other evidence merely to contradict the denial.
• “Fishing expedition”: A pejorative term for asking questions without a factual basis in the hope of uncovering something favourable. Courts prevent this to protect fairness and avoid prejudice.
• Foundation for cross-examination: Before a party can question a witness about a third-party communication (e.g., a social media message), they must establish a connection between the witness and the communication. Without that link, the topic is off-limits.
• Attribution: In the context of digital messages, attribution asks, “Who sent this?” If a complainant admits to operating an account or sending a message, that admission opens up potential cross-examination on content; if not, the avenue may be closed.
• Summing-up discretion: Trial judges decide how to frame and summarise evidence and issues for the jury, provided they do so fairly and accurately. They need not refer to every point, especially material not in evidence before the jury.
• Complainant anonymity: Under the Sexual Offences (Amendment) Act 1992, complainants in sexual offence cases have lifelong anonymity. Courts and commentators avoid identifying details to prevent “jigsaw” identification.

Practical Guidance for Practitioners

• Defence:
  • When relying on third-party digital messages, separate your application into (a) attribution questions and (b) content questions. Make it explicit that you seek permission to ask narrow attribution questions first.
  • Record and, if needed, renew requests that the prosecution put attribution to the witness. Preserve this issue for appeal if not actioned.
  • Recognise the “bound by the answer” rule—plan strategy accordingly.
• Prosecution:
  • Act on reasonable, focused defence requests to ask witnesses attribution questions about digital communications. Document the steps taken and the answers obtained.
  • Be alert to the difference between speculative forays and targeted inquiries whose answers could affect trial fairness.
• Judges:
  • Maintain the prohibition on fishing expeditions while facilitating narrow, foundational attribution questioning where appropriate. This balances fairness to the complainant with the accused’s right to test credibility.
  • In summings-up, focus on the live issues and evidence actually before the jury; avoid comment on materials not in evidence, such as unadduced portions of witness statements.

Conclusion

R v Yu clarifies two interlinked trial principles in the context of digital communications and sexual offence prosecutions. First, the court reaffirmed that cross-examination premised on third-party messages requires a proper evidential foundation—without attribution, it will be halted as a fishing expedition. Second, the court gave procedural teeth to the prosecution’s obligation to pursue reasonable, defence-flagged attribution inquiries: when simple, targeted questions could unlock a legitimate credibility line, they should be asked.

The Court’s case management directions underscore that the answer to those attribution questions can be decisive. A denial likely closes the issue; an admission may render a conviction unsafe. By drawing this practical line and insisting on follow-through, the Court has provided a roadmap for handling digital-messaging credibility disputes—one that promotes fairness, procedural clarity, and efficient resolution. The judgment also reaffirms judicial discretion in summings-up and the limits of collateral credit challenges.

In sum, Yu stands as a timely guide for trials in the digital era: lay the foundation, ask the right questions, and do not neglect simple investigative steps that may matter greatly to the jury’s assessment of credibility and the safety of the verdict.