A security service is entrusted with secrets because some public work cannot be conducted in full view. That grant of confidence is considerable, but it is not unlimited. When an agency supplies a court with evidence based on falsehoods, the central question is larger than whether one statement was inaccurate. Citizens must know whether the institutions charged with testing evidence can still perform that duty when secrecy is invoked.

BBC News reports that MI5 has accepted that it gave evidence based on lies in a neo-Nazi spy case while defending a violent informant. The supplied account says this is the first time the agency has confirmed the BBC's revelation. Those facts alone warrant careful public attention. They do not, however, justify speculation about undisclosed evidence, individual motives, or matters beyond the report.

The first task is to define the failure

Official phrases can make a grave error sound foggy. Evidence based on lies is not the same thing as a disputed interpretation, an incomplete recollection, or a clerical mistake. Any formal review should identify, in plain terms, which proposition was false, who knew what at each stage, and when the institution became capable of correcting the record.

This need not require publication of operational details. A court can protect identities and sensitive methods while still demanding an exact account of how unreliable material entered the proceeding. Secrecy may govern the contents of intelligence. It should not erase the chain of responsibility by which that intelligence becomes evidence.

The distinction matters because remedies depend on causes. A false statement repeated unknowingly points toward verification failures. A warning ignored by supervisors points toward management and escalation failures. A statement maintained after its falsity became known raises a still more serious question about institutional candor. The public need not guess which occurred. The responsible authorities should establish it from records.

Correction must reach every affected decision

Correcting a sentence in one file may be necessary, but it is not necessarily sufficient. False evidence can influence later judgments, internal assessments, warrants, restrictions, or accounts given to oversight bodies. A proper review should therefore follow the information outward. Where did it travel? Who relied upon it? Was it repeated in another form? Did any decision remain in force after the underlying claim became doubtful?

This is ordinary record keeping applied to an extraordinary institution. Agencies routinely track the handling of sensitive information. They should be equally capable of tracing the distribution of a material error. Every affected recipient should receive a correction that is as clear and prominent as the original assertion.

The court itself also requires more than reassurance. Judges depend upon parties to distinguish verified fact, intelligence assessment, hearsay, and uncertainty. When national security limits what can be tested openly, the duty of candor becomes more important, not less. The party asking for secrecy also assumes a heavier obligation to describe the concealed material faithfully.

Oversight should test the system, not merely the apology

An institutional admission is a beginning. Effective oversight asks whether procedures now make a repetition less likely. Reviewers should examine how informant claims are corroborated, how doubts are recorded, who may halt a submission, and how corrections reach courts after filing. They should also determine whether staff can report concerns beyond the immediate chain of command without losing the substance of the warning along the way.

Useful public reporting would answer a modest set of questions without disclosing protected operations. Was the inaccurate material identified? Were all proceedings touched by it reviewed? Were responsible bodies notified? Have evidentiary safeguards changed? Is there a schedule for reporting completion? These are administrative questions, but constitutional government often rests upon sound administration.

There is a temptation in intelligence controversies to choose between two absolutes: disclose everything or trust the agency entirely. A durable democracy accepts neither. It permits legitimate secrecy while insisting upon independent scrutiny, documented correction, and consequences proportionate to the findings.

The public does not need theatrical contrition. It needs a repaired record and proof that the repair extends wherever the false evidence went. Security institutions preserve authority when they show that secrecy is a disciplined public trust, not a shelter from accountability.