and
November 4, 2024   12 mins

BBC Verify, the new fact-checking service inside the BBC, was launched last year with much fanfare and at great expense. Its verdicts are pushed across all the corporation’s channels. So why is it getting so many things wrong? And where is the accountability for these arbiters of truth? UnHerd Investigations Editor, David Rose, subjects Verify to some fact-checking of his own…

On 6 August, at the height of the riots that swept Britain after the Southport murders, BBC Verify revealed “one of the most shocking” outbreaks of violence: a racist attack on Humberside. According to Verify, “an angry mob of white men surrounded and attacked a car with men of Asian heritage inside.” Some of the attackers “threw bricks and shouted racist slurs. One performed the Nazi salute.”

The Verify team deployed its OSINT (open source intelligence) skills and facial recognition software to compare images from the scene with social media profiles. Soon enough, they’d identified and named two of the alleged attackers. Yet despite having no less than four writers on the story — just one of many Verify articles to boast multiple bylines — this gaggle of reporters initially failed to spot the most important fact of all. 

That only became clear later, in a note added at the bottom of the BBC website once the story had been published. “An earlier version of this article identified the men in the car as of Asian heritage,” it said. “Humberside Police has since told the BBC they are Eastern European.” To put it differently, the victims on Humberside were as white as their attackers. Whatever else the incident was, then, it certainly wasn’t a straightforward case of white racism against non-white targets. Later, the BBC also changed the headline.

BBC Verify, the new fact-checking service inside the BBC, was launched last year with much fanfare and at great expense. Its verdicts are pushed across all the corporation’s channels. So why is it getting so many things wrong? And where is the accountability for these arbiters of truth? UnHerd Investigations Editor, David Rose, subjects Verify to some fact-checking of his own…

On 6 August, at the height of the riots that swept Britain after the Southport murders, BBC Verify revealed “one of the most shocking” outbreaks of violence: a racist attack on Humberside. According to Verify, “an angry mob of white men surrounded and attacked a car with men of Asian heritage inside.” Some of the attackers “threw bricks and shouted racist slurs. One performed the Nazi salute.”

The Verify team deployed its OSINT (open source intelligence) skills and facial recognition software to compare images from the scene with social media profiles. Soon enough, they’d identified and named two of the alleged attackers. Yet despite having no less than four writers on the story — just one of many Verify articles to boast multiple bylines — this gaggle of reporters initially failed to spot the most important fact of all. 

That only became clear later, in a note added at the bottom of the BBC website once the story had been published. “An earlier version of this article identified the men in the car as of Asian heritage,” it said. “Humberside Police has since told the BBC they are Eastern European.” To put it differently, the victims on Humberside were as white as their attackers. Whatever else the incident was, then, it certainly wasn’t a straightforward case of white racism against non-white targets. Later, the BBC also changed the headline.

If journalism is the first draft of history, often produced at speed, it’s inevitable that reporters sometimes get things wrong. But just as the 6 August story suffered from a glaring and arguably story-ending error, BBC Verify seems to suffer more mistakes than most. Having examined more than 200 stories listed on BBC Verify’s website, and produced between its launch in May 2023 and 5 October this year, we found 12 that were later corrected, clarified or withdrawn. That’s about one in 20.

For an outfit like BBC Verify, an expensive new unit established to combat disinformation, that’s awkward enough. That’s doubly true when journalists elsewhere in the corporation face redundancy, with £24 million annual cuts to the BBC News budget forcing executives to axe well-regarded programmes and shed 130 staff. Yet beyond these individual embarrassments, or the strange focus on Verify to the detriment of other reporting, the real story here is what BBC Verify says about the contemporary BBC’s approach to journalism — and how it often seems to put shallow, memeable newsgathering ahead of the facts. 

BBC Verify was born in an age of disinformation, in which “fake news” spread by platforms such as TikTok has eroded trust in “legacy media” — and diminished its audience. A high-level BBC insider says that from the moment Deborah Turness was appointed chief executive of BBC News in 2022, she tried to persuade her senior colleagues that the way to restore both was to set up a unit to challenge the spread of disinformation and report the facts. She got her way, and BBC Verify was the result. 

Despite being founded around 18 months ago, Verify has quickly come to dominate BBC News output. Boasting over 60 dedicated journalists, it enjoys a special corner of the BBC website, with each story it publishes highlighted by a BBC Verify logo. Covering everything from war (“What satellite images reveal about Israel’s strikes on Iran”) to climate change (“No, Hurricane Milton was not ‘engineered’”), online content is dovetailed by Verified Live, which airs weekdays on BBC News. No less striking, Verify has also made stars of several BBC journalists. One example is Ros Atkins, the BBC’s analysis editor. Another is Marianna Spring, a disinformation correspondent who’s appeared everywhere from Breakfast to Radio 4, and so become one of the corporation’s most recognisable faces.

Given all this coverage, at any rate, it’s unsurprising that Verify should also enjoy praise from executives. “BBC Verify is transparency in action fact-checking, verifying video, countering disinformation, analysing data and explaining complex stories in the pursuit of truth,” is how Turness put it. As Verify’s creator has added, its staff aren’t just ordinary journalists. Rather, they’re “a highly specialised team with a range of forensic and OSINT capabilities that enable them to go beyond conventional newsroom techniques”. 

Last month Turness sent a message to every BBC journalist making further grandiose claims. In the note, seen by UnHerd, she said Verify was “on the front line” of the corporation’s “fight for truth”, and its “forensic teams” were “setting the standard for transparent journalism”. Such enthusiasm is clearly echoed in the accounting department too: the BBC has said the salary bill for Verify’s 63 staff is £3.2 million a year.

Yet amid the plaudits and the paychecks, just how accurate is Verify’s reporting? 

Certainly, that the unit has had to correct, clarify or withdraw about one in 20 stories isn’t a good start. “The BBC claims that BBC Verify is on the frontline of the ‘fight for truth’ but the fact it has been forced to correct 5% of its stories suggests otherwise,” explains Danny Cohen, the former director of BBC Television. “That is a very high number for any news organisation, and the BBC should be upholding the highest of standards.” 

“The BBC claims that BBC Verify is on the frontline of the ‘fight for truth’ — but the fact it has been forced to correct 5% of its stories suggests otherwise”

Nor do things look much better once you dig a little deeper. Like the Humberside story, after all, some mistakes involve that most sensitive of subjects: race. On 25 February, for instance, BBC Verify published a troubling exposé. Its author, Maryam Ahmed, had used advanced statistical techniques to compare thousands of car insurance quotes, finding that drivers in racially diverse areas were having to pay much higher premiums an average of 33% more than people from places that were mainly white.

It seemed the motor insurance industry was racist, for behind the disparity lay what the BBC called an “ethnicity penalty”. Even if differing levels of accidents, crime and economic prosperity were taken into account, “areas with a high number of people from ethnic minorities saw higher prices”, the article said. It went on to quote a demand from Citizens Advice that the Financial Conduct Authority “get off the sidelines and investigate why people of colour are being charged so much more”.

Seven months later, the BBC quietly replaced the article with an embarrassing statement. It now transpired there was no “ethnicity penalty” after all. The BBC “had not established why premiums were higher”, because it had failed to see “the limitations of the data we used”. The story “did not meet our normal editorial standards” and was therefore being deleted.

And if that’s echoed by errors in other domestic stories — among other things, that includes a muddle over high speed railways — the unit has arguably faced its biggest problems very far from Britain. Since the Hamas attack of October 2023, indeed, the Middle East has proved something of a theme for the unit: our analysis shows that in the year since the attack, about a fifth of the stories listed on its website have focused on the region. 

And if Cohen says this suggests an “unhealthy obsession” with events in Israel, some of Verify’s claims on this most controversial of conflicts don’t stand up.

Early in the war, on 17 October last year, an explosion at the Al Ahli hospital in Gaza triggered furious controversy, after the BBC correspondent Jon Donnison said on air that it was “hard to see” what could have caused it other than an Israeli airstrike. 

Israel insisted that the blast was caused by a Palestinian rocket, but according to Donnison, these might kill “half a dozen maybe a few more”, and “when we’ve seen rockets being fired out of Gaza, we never see explosions of this scale”. In its reports that evening, the BBC echoed a claim by Hamas that it had killed 500 people. In fact, the hospital itself had not been hit: the blast was in the car park outside, and the death toll, according to both Western governments and Human Rights Watch, was far lower than 500.  

BBC Verify published three successive items on the event, culminating in a lengthy article on 26 October that carried the bylines of no fewer than 12 of its staff. 

By this time, the Biden administration and numerous independent experts had concluded that the explosion was caused by a misfired rocket, launched from inside Gaza. Rishi Sunak, the then-prime minister, had told Parliament this was likely true, based on the “deep knowledge and analysis of our intelligence and weapons experts”. Yet BBC Verify continued to maintain the cause of the blast was still “contested and that, as Hamas still claims, it might have been an Israeli strike. 

Verify’s assertion relied on analysis by an organisation called Forensic Architecture, which the BBC somewhat obliquely described as a “UK-based organisation which investigates human rights abuses”. 

What it failed to mention was that Forensic Architecture’s advisory board includes Ryvka Barnard, the deputy director of the Palestine Solidarity Campaign; and that it has a long record of issuing reports critical of Israel. At the same time, Forensic Architecture has often worked with Al Haq, a Palestinian group designated a terrorist organisation by Israel. After the October 7 attacks, meanwhile, Al Haq staff issued numerous statements describing them as legitimate acts of resistance. One of its leaders urged his social media followers to support the “jihad”. 

Despite all this, Verify’s initial story is still online — complete with the claim that the origin of the hospital blast remains “contested”. That’s even as the story now includes a clarification at the bottom conceding that Forensic Architecture itself now admits “they had been wrong” about the specifics of the attack.

Nor is this Verify’s only Gaza story to be based on dubious sources. On 1 March, it published a report into the deaths of 112 Gazans, who’d died the previous day while food was being distributed from a convoy of aid trucks. Hamas claimed most of the dead had been shot by the IDF; Israel that they’d been crushed in a stampede.

The six named Verify staff accredited in its report analysed satellite photos, videos and drone footage. It seems OSINT wasn’t much help, as they still weren’t able to draw any definitive conclusions. In the end, Verify relied on an old-fashioned human source. A Palestinian named Mahmoud Awadeyah told the reporters he’d seen what happened: a deliberate attack on desperate civilians by Israeli soldiers.

Beyond Awadeyah’s testimony, Verify’s story contained no other evidence this was true. More to the point, other research suggests that he may not be a reliable source for such serious allegations. For one thing, he has worked for Tasnim, a news agency controlled by Iran’s Revolutionary Guards. On the same day as the murder of seven Jews outside a Jerusalem synagogue, meanwhile, Awadeyah tweeted about “A state of rejoicing, exuberance, and mosques confirmed with exuberance.” The same message included an emoji of a contented-looking face with a halo. He’s also posted photos of himself with Khalil Al Bahtini, a leader of the military wing of Palestinian Islamic Jihad, and mourned his “sincerity and loyalty” when the latter was killed in an Israeli airstrike.

Twelve days after the Verify piece was published, the BBC appended yet another “clarification” — noting that “this piece has been amended to make clear Mahmoud Awadeyah’s current employment” for a Lebanese news station called Al Mayadeen. For its part, Verify also noted Al Mayadeen’s sympathies to groups fighting Israel. But even that’s not the full story: Al Mayadeen openly supports Iran, Hezbollah and the murderous Assad regime in Syria. 

Cohen, for his part, is amazed that the BBC didn’t feel the need to explain that an eyewitness source was employed by a pro-Iranian outlet. “This is a good example of bias by omission,” he says. “Without all the facts and context clearly presented, the BBC is misleading audiences.”

And if Verify has been caught out by its sources in other areas too — an expert quoted in support of a wealth tax has campaigned for just such a policy, even as his background wasn’t made clear — reporting on riots or war at least feels worthy in principle. 

Elsewhere, though, the unit seems content paddling in distinctly shallow waters. On 7 May, for instance, Spring presented a filmed segment describing a fake picture on Instagram. This was significant, she said, because AI-generated forgeries often involved politicians, and so could potentially influence elections. In the event, though, the picture she highlighted wasn’t of Donald Trump or Joe Biden — but rather of the singer Katy Perry, doing nothing more than standing in a floral dress she hadn’t actually worn. Undaunted, Spring justified her prime time report by saying it was “a reminder to put our social media sleuthing to the test”.

Nor is this the only story that feels like it could have been left to the tabloids. On 17 October, BBC Verify published a story about a supposed “wave of cases” involving an alleged drug called Devil’s Breath. The article explained that it was apparently a travel sickness remedy, which could “be administered to victims in the street, without them realising they have been drugged”. Perhaps it can. Yet Verify admitted there was “no evidence” it had been administered in the cases it examined, though it “has been used in robberies in Ecuador, France and Vietnam”. Even if it had been used in Britain, meanwhile, this would have been “difficult to establish”.

Sometimes, Verify even “proves” things that no one seriously disputes. One example came on 7 September, when it published images showing Israel had been building a new road along the border between Gaza and Egypt. Three weeks later, it analysed videos of the airstrike that killed Hassan Nasrallah in Beirut. The fact that solar panels were seen facing away from the bombed building suggested that the attack had come from the south: in other words from Israel. Who knew?

Combined with those monumental bylines — even a straightforward story about flooding in Libya involved a dozen staff — and that £3.2 million wage bill begins to feel oddly high. Indeed, Verify might cost the BBC even more than that. Notwithstanding the unit’s focus on transparency, when we asked the corporation about the rest of its budget, for costs such as travel and OSINT software, a spokeswoman said they weren’t available because they’re “accounted for as part of the overall BBC News budget”. 

Such munificence might matter less if the BBC was rolling in money. As it is, the corporation is seriously cash-strapped, a situation that’s obliged managers to cancel some of its most respected shows. Newsnight, for one, is now a shadow of its former self, shorn of filmed reports and investigations. HARDtalk, an in-depth interview programme, is vanishing too, despite first going on air back in 1997. Together with those 130 redundancies, and the high-level BBC insider is unsurprisingly upset about the corporation’s apparent focus on BBC Verify, especially when they suspect it’s indulged as Deborah Turness’s baby. “Because she is the head of news, it can only ever be declared a success,” he explains. “Its triumph is pre-ordained.” It hardly helps, the insider adds, that other sections of BBC News actually have to correct stories far less often than Verify — even as they “they don’t presume to be more truthful than anyone else”. 

Nor do things seem likely to change going forward. In late October, mere days after BBC director-general Tim Davie made a speech about the growing threat of “misinformation” from foreign powers, the corporation had ordered four of its most seasoned foreign correspondents back to Britain to save money. All had years of experience, backed by foreign language skills, of reporting on authoritarian regimes including China, Russia and North Korea. Now, though, they’d been told to cover news in Britain.

Within the BBC, the news was received with dismay. The Times quoted a staffer saying they had been “told to return at a time when the amount of propaganda we are up against is astounding. None of them costs very much money and strategically, it doesn’t make sense when you look at the big, looming stories.” But in the new BBC, it seems that on-the-ground reporting, battling to get at the truth, is destined to be replaced by analysis of videos shot by amateurs on mobile phones — all carried out by BBC Verify in London.

Certainly, the corporation clearly remains optimistic about the unit. 

We asked the BBC to comment on everything in this article. Its spokeswoman told us: “By their very nature, BBC Verify stories are heavy in facts and detail on contentious topics which can sometimes require clarifications and updates, and occasionally corrections, which we do in an openly transparent way. Often pieces are updated to show our workings to audiences — showing them how stories develop as new information comes to light. We do not recognise the flawed calculations that attempt to describe updated pieces with transparent workings as corrected articles.

“We are proud of the vital work that BBC Verify is doing, and our audience research shows that people value this offering and are actively seeking out BBC Verify stories to help them understand events.”

Regarding the false claim that the victims of the attack on the car during the riots were of Asian heritage, the spokesperson said: “The story used a range of technology, on the ground reports and social media to piece together how the violence escalated in Humberside that day. When we became aware of the new information we amended the story swiftly and transparently. We stand by the story as an accurate account of what happened.”

As for the now-deleted article about “racist” car insurance, the spokeswoman said that “we have been transparent in saying this story did not meet our normal editorial guidelines and why we withdrew the article”. She also defended the report claiming Israel had massacred Gazans when they tried to get food, despite relying on a single source who’d worked for a pro-Iran TV channel. “The fact,” she said, “that someone has expressed an opinion on social media doesn’t automatically disqualify them from giving eye-witness testimony.” 

Not everyone at the corporation is so bullish. Our BBC insider was unimpressed by these comments, saying he was dismayed that even when Verify had admitted that fundamental elements of major stories were wrong, they’d nonetheless stayed online. As he puts it: “This is misleading and dishonest.” 

Beyond the finances, questions about the unit surely remain. Scanning our analysis of the stories on BBC Verify’s website, it’s hard to avoid the impression that it’s selective with its targets: evidence, perhaps, of a certain Left-leaning bias. It doesn’t, for instance, examine claims made by Hamas with anything like the frequency and rigour it applies to Israel. During the UK election, meanwhile, it dismissed Reform UK’s manifesto by saying its promises had been described as impossible by “experts”, “economists” and “analysts” — without naming or quoting any of them. As was noted last week, meanwhile, it failed to measure the Budget, and the tax increases it contained, against Labour’s pledges during the election. 

Especially when the corporation is tightening its belt elsewhere, it’s little wonder Cohen remains sceptical. “At a time when the BBC is having to cut jobs and save a lot of money,” he concludes, “the track record of BBC Verify seems very hard to justify.”


David Rose is UnHerd‘s Investigations Editor.

DavidRoseUK