News

AP launches verification dashboard for publishers to meet ‘demand for authenticity’

Press Gazette · Alice Brooker · last updated

Associated Press has launched an AI-powered tool to help journalists verify text, photos and videos in one place.

The AP Verify dashboard will let subscribers access several verification tools on one centralised platform.

It includes an AI chatbot assistant, geolocation, object and landmark recognition, AI text detection, transcription services and an anti-duplication tool to alert users if colleagues are verifying the same content.

The aim is to help journalists verify content purporting to show major news events in a world where AI is making it harder to tell what’s real and what’s not.

AP Verify has been used by AP journalists for the past year and is now being offered to other publishers.

Aimee Rinehart, senior product manager AI strategy for The Associated Press, told Press Gazette that AP used the tool to successfully secure original Texas flood footage by using reverse search and people finder tools to contact the videographer and secure permission.

The newsroom also found a key eyewitness video from the Charlie Kirk assassination by analysing social content and using people search tools to confirm the source. 

Rinehart said: “It’s a natural extension that that AP would develop a tool like AP Verify so that everybody else can make sure that their information streams are also producing and sharing out authentic images, authentic videos, and being able to check that origin story of information.”

Each newsroom with a subscription has their content kept private, so publishers cannot see what their competitors are verifying.

Before the dashboard was introduced, AP journalists were using a “patchwork of tools” to help verify information, including Google reverse image search, and this was one of the reasons for wanting a “comprehensive dashboard”, said Rinehart.

She added that knowing a publisher has uploaded and verified content through the dashboard would give the public “another level of confidence” in its output.

“No one should be clearing stories or writing stories without understanding how information travels online,” said Rinehart.

The dashboard relies on several third-party providers to help verify content, including Google’s Fact Check and image reader, Trint for transcription, Graylark’s geolocation, GPTZero’s generative-AI text detection, Trendolizer, and identity solutions provider Pipl.

“None of the tools are 100%… we would never recommend you go straight to publish just based on that tool’s information,” said Rinehart.

These tools “only work if [used] side-by-side with a human”, and are used as a “guide” – “sometimes there are false flags”, she added.

Rinehart said there are “a lot of people, particularly in small, under-resourced newsrooms, that are surprised that you don’t just have to wing it when it comes to publishing things”.

“Even when we did market research for this, there was a local broadcaster who said, ‘Yeah, sometimes we’ve run the wrong tornado video and then we apologise the next day’,” she said. “It doesn’t have to be that way.”

AP was ‘source of truth’ on Kate Middleton photo ‘manipulation’

Rinehart said the launch of the dashboard will “test” if there’s a demand in the publishing industry for the centralisation of these tools.

“We want everyone to think that they too can verify online information, because it really should be as accessible to people as that,” she said.

In 2024, five international news agencies including AP removed a photo issued by Kensington Palace of Kate, Princess of Wales, and her children over accusations it had been “manipulated”.

“I think what’s in demand from publishers, and we saw this with the Middleton example, is that AP became a source of truth for so many publishers when we retracted that image,” Rinehart said.

“That’s what we really want to get to, is everybody trying to discern what is real, what has been retouched, and what can we trust online?

“I don’t know if everybody’s going to be banging down the doors for AP Verify… but I do think the demand for authenticity is there.”

The post AP launches verification dashboard for publishers to meet ‘demand for authenticity’ appeared first on Press Gazette.

Related Knowledge Center: Artificial intelligence

State of Play

On top of a failing business model, newsrooms must navigate the powerful — and not fully reliable — new capabilities of artificial intelligence and the deployment of large language models.

State of the business model | Copyright and intellectual property


More informationabout Artificial intelligence

Related Knowledge Center: Information integrity

State of Play

In an increasingly fractured media landscape, truth has been a casualty in the public discussion. Coverage has been warped by the business needs of struggling news organizations, and a significant segment of the public relies on bad-faith outlets that are little more than propagandists. While fact-checking has become routine, some organizations are exploring other ways to combat the trends. And misinformation spewed by AI-driven chatbots has become more of a feature than simply a bug. 

Fact-checking and its detractors


More informationabout Information integrity

Related Knowledge Center: Inside the newsroom

State of Play

As newsrooms struggle with rickety business models and precarious finances — as well as new technologies —  they are deploying an array of editorial responses. 

The Times and its critics  |  Washington Post mortem    


More informationabout Inside the newsroom