Interview

April 9, 2021

How Logically uses AI — and humans — to tackle misinformation

Most fact-checkers rely on humans — Logically uses AI to speed up the process, allowing it to execute at scale.


Karam Filfilan

6 min read

Sponsored by

Fake news, misinformation, alternative facts: it’s becoming increasingly difficult to separate fact from fiction when it comes to what we read online. 

That’s why Yorkshire-based startup Logically has created a series of AI-based products to help people, businesses and governments identify false stories and narratives on the internet, stop their spread and track regular sources of misinformation. 

Logically’s success so far has led it to identify 130k fake stories on India’s 2019 election and raise £2.5m in investment to work with social media giants covering the recent US elections. It even debunked one of the sources of the widespread conspiracy theory that Covid-19 is a plot to destroy human society with 5G.

Advertisement

Sifted spoke to CEO and founder Lyric Jain about how it combats fake news and why Yorkshire’s proximity to universities made it the perfect place to build Logically. 

Fast fact-checking

Founded in 2017 in West Yorkshire, Logically uses a team of more than 100 global data scientists, coders, writers and developers. Along with AI modelling, the startup analyses data from over a million publicly available media sources and social media platforms to assess news story credibility and the spread of false narratives.

After the launch of our consumer products, we spent the next couple of years working on how we could create solutions that tackle the information crisis at scale.

Logically works via a free mobile app which allows users to verify content by sharing a link with the app. It also developed a browser extension that fact checks news sites and social media as users surf the internet. Available in the US, UK and India, more than 200k people have downloaded Logically’s customer facing products. But Jain says the team always felt they could operate on a much larger scale thanks to advanced technology.

“After the launch of our consumer products, we spent the next couple of years working on how we could create solutions that tackle the information crisis at scale. We developed state-of-the-art AI and advanced natural language processing (NLP) that can detect problematic content online, track disinformation campaigns and help to coordinate countermeasures for governments, platforms and businesses worldwide,” he says. 

This culminated in the release of Logically Intelligence in March, a threat intelligence platform for use by public bodies and governments to detect and analyse online threats and disinformation. 

Logically's new threat intelligence platform, Logically Intelligence.

Humans and AI in partnership

Logically’s use of AI sets it apart from other fact-checking businesses, which typically rely on humans to sift through information to determine a story’s validity — like Snopes, a US-based fact-checking platform which previously partnered with Facebook to monitor the site’s content. They had to pull out of the partnership because it was “impossible to keep up,” Snopes VP Vinny Green told Poynter. Logically’s reliance on AI speeds up the process, allowing it to execute at scale.

“Our AI models identify claims and rank them based on credibility and accuracy, as well as cross-referencing data related to the claim before it’s manually verified. If claims have been checked before, a result can be given almost instantly,” he says. 

Newer or more complex claims go through the system but are flagged for review by human moderators. But Jain said in an interview with Wired that in the future, Logically aims to automate as much of the process as possible, making fact-checking much quicker. 

Our AI models identify claims and rank them based on credibility and accuracy, as well as cross-referencing data related to the claim before it’s manually verified. If claims have been checked before, a result can be given almost instantly.

Ultimately, human involvement will always be necessary — and it starts with the person sharing the article in the first place. “Everyone has a part to play in slowing the spread of fake news and misinformation. We should challenge ourselves to think about the information we’re sharing before we do so, especially if it seems controversial or sensational,” says Jain. 

Northern powerhouse

Jain founded Logically with an £8,250 startup grant from the Massachusetts Institute of Technology (MIT), where he studied engineering and computer science. He chose to base the company in Brighouse near Leeds; he had previously worked in the region, so already had a network of local business contacts. 

Advertisement

“Leeds has made many great strides over the past few years to become a thriving northern tech hub and I wanted Logically to be part of that,” he says. 

“One of the main aspects that appeals is the proximity to so many universities which excel in data science and machine learning. Being able to consistently access a high quality talent base due to centres of excellence like the Universities of Leeds and Sheffield is critical to our business,” he adds.

Logically CEO and founder Lyric Jain

Jain also cites the local partnerships Logically has fostered in Yorkshire, particularly with larger businesses who provide mentoring and networking events. Logically was part of PwC’s Scale Yorkshire programme in 2019, a 10-week series of pitches, networking events and training that allowed him to meet likeminded entrepreneurs and open new networks. 

During the Google for Startups UK Immersion in 2020, Logically was connected to fellow founders from across the UK going through similar scaling challenges. The programme focused on helping founders upskill themselves, their leadership teams and learning best practices on everything from OKRs to hiring at scale.

Logically also entered into Knowledge Transfer Partnerships with local universities through Innovate UK, which Jain says have been “incredibly successful at accelerating technical research and developing better end products.” These work by partnering a business with a ‘knowledge base’ — like a university or research institution — and a graduate who leads on a specific business project. The business gains academic expertise it might not have in-house, and the graduate gets a paid position that utilises their expertise. 

Jain believes Yorkshire’s growing tech scene means that more businesses will move to the region — and that creative talent will more likely stay in Yorkshire instead of joining the brain drain to London.

We’ve seen a trend of scaleups moving here from London, as well as people happy to move away from London for a role.

“We’ve seen a trend of scaleups moving here from London, as well as people happy to move away from London for a role. That’s partly a result of the growing tech hub and supportive ecosystem in Leeds, but also large, creative organisations like Channel 4 and Sky moving substantial teams here,” he says.

“Everything I’ve seen over the past few years has validated my decision to start and build Logically in Yorkshire. Yes, having good links with London is important, but everything you need is here,” says Jain. 

This is the second article in our ‘Local Heroes’ series, where we explore ecosystems and startup communities outside of London through the eyes of founders. You can read the first in our series here, which is about how Edinburgh’s Neatebox builds apps to help disabled people live independently.

Find out more about the Google for Startups UK Immersion programme and startups here.