The case of Joanne Fraill, who faces jail for breaching jury rules by contacting a defendant through Facebook, is a reminder of the seriousness of the challenge to the British justice system posed by the internet. There is more at stake here than Tweeting about the private lives of celebrities.

It is no exaggeration to say that trial by jury, that pillar of the justice system, could become impossible if some way is not found to prevent jurors abusing the internet during cases. Juries will be susceptible to prejudice from reports and gossip they find online, evidence ruled as inadmissible by the court may be easily accessed, and past convictions — normally not disclosed — may be traced. The result will be collapsing trials and unsafe verdicts.

If anybody doubted whether jurors were reading this sort of material, an excellent investigation this week by the Times (sadly behind its paywall) has put the matter to rest, their trawl finding 40 examples of apparently illegal postings and online statements by jurors, including several who declared defendants guilty before hearing all the evidence and one who conducted an online poll of his friends on what his verdict should be.

It is easy to understand jurors googling cases they have been assigned to or even Tweeting about events in court, but the legal system as presently constituted simply can’t survive if it is anything more than a rarity. Anyone found guilty in a case tainted by jury misconduct is likely to have their conviction overturned, for the good reason that they can argue they did not have a fair trial.

I covered the case of Barry George, who was convicted and much later cleared of killing Jill Dando in 1999. I always believed that the jury at his first trial was influenced by the monstering he had received in the press at the time of his arrest, even though that had been a year earlier. Others argued that in that case and most others, the time between charging and trial was sufficient for jurors’ memories of prejudicial coverage to fade. That argument can no longer be said to apply, because if jurors are googling defendants’ names, there is no fade factor. What pops up on their screens is exactly what was written a year earlier, and it can be highly prejudicial.

Then there is the problem of previous convictions. It is axiomatic in British law that jurors try a defendant on the evidence in the case alone; they are not normally told about previous convictions in case they succumb to the temptation to convict someone “because he has done that sort of thing before”. But if they can easily find a record of those convictions on their smartphones the court’s aim is subverted and the defendant does not receive a fair trial.

Those are just two problems of many. Will it be enough for judges to warn jurors more firmly that they must not consult the internet? I doubt it. As things stand it is just too easy to do and too hard to detect. I suspect before long we will see jury consultations being officially monitored to ensure no one introduces improperly acquired information. That, however, is unlikely to be enough on its own to prevent tainted verdicts. In time either the courts or jury members or the internet will have to change. Which will it be?

Brian Cathcart teaches journalism at Kingston University London and tweets at @briancathcart