IFCN: “If AI is Used by Humans, it Must Also be Supervised by Humans”

We have gone from a timeline in which the relevance of traditional media gave way to social media and online networks, replacing it in dominance and global dissemination. Nowadays, it is quite difficult to read something first in a physical newspaper than on a website! On the other hand, we are increasingly witnessing a systematic war on reliable and impartial information, with the discrediting of sources, where opinion has prevailed over facts and entertainment over real news.

The Poynter Institute’s International Fact-Checking Network (IFCN) was launched in 2015 to bring together the growing community of fact-checkers around the world. The network defends information integrity in the global fight against disinformation, reaching more than 170 fact-checking organisations through advocacy, training and global events. Director Angie Holan spoke to our HQ Magazine Manager, Manuel Fernandes, about the concepts of infotainment, disinformation, truth and lies.

Headquarters: We have moved from a timeline in which the preponderance of traditional media was fundamental, to one in which digital media and social media outreach have replaced it in dominance and global dissemination. How do you see the role of the ‘editor’ or ‘proofreader’ in this online age? How can the ability to check facts and statements be transferred from paper to online?

Angie Holan: It is critical that we transfer what we now consider traditional print competences to the online space. They may be different in the way people interact with the various media, but they are the same in the sense that factual standards need to be maintained in both areas. So, the skills that fact-checkers bring to verifying traditional media and modern media are the same, with an emphasis on primary sources, documentation, expertise, etc. In a way, we need to make these old methods of processing information relevant and dynamic for an online area. For example, today we see fact-checkers who write lengthy reports turning them into short chat videos on TikTok.

What we need to ensure is that when we present information as validated, verified and accurate, it follows the standards of evidence. A big part of my organisation's role is to maintain standards of evidence among fact-checkers. We can go into this in a lot of detail, but I think the important thing to note is that there are standards of verification, and some people abide by them and others do not. Often, you have to spend a bit more time finding out whether your information is really verified or not. Whenever there is a question about the integrity of information, it is necessary to be able to investigate the evidence so that it can be reproduced. This applies to both online and offline. The problem is that the volume and speed of information online is so fast that people often do not slow down to check. That is where fact-checkers come in. We are slowing down to verify and showcase our work in various media including online platforms. Right now, the current climate points in that direction.
 

HQ: I don’t think it’s an exaggeration to say that we are living through a moment in which reliable, impartial information is under systematic attack, as trusted sources are increasingly discredited and unchecked opinion and entertainment too often eclipse facts and serious reporting. How do you view these new forms of infotainment and ‘community notes’ that certain media organisations and technology platforms are pushing?

AH: One of the problems we face with these technologies is that we have not yet developed new signifiers of high-quality information that users can view and understand quickly. In many contexts, people really need to do their homework and check for themselves whether what they consume is verified or not. This can be very time-consuming, cumbersome and inefficient. We have a serious problem with information integrity today, because there are no efficient ways of doing this. However, I believe that, as a society, we will develop systems to deal with this vast amount of information more efficiently and comfortably. It is a very uncomfortable experience these days to go online and know that everything you see may or may not be true. We also have to think about the question of what expertise is. For specialists, there has never been so much information available as there is now, which leads to more in-depth research. But for non-experts, the environment is very challenging.
 

“I think there are some very deep problems with the business model for high-quality content in order to attract audiences”

HQ: Further, several websites and news outlets have developed real-time fact-checking tools to counter disinformation and restore accuracy to public debate. This work is especially vital at a time when ‘truth’ can be twisted into ‘lies’ through media spin, attacks on the credibility of institutions, and the persistent misuse of the dubious concept of ‘fake news’. What tools and resources can public institutions, associations and media apply in their activities to preserve their reputation and image?

AH: As a foundation, I believe we need strong education systems. That is where young people learn to read, write, do maths, science, think critically and analyse arguments. A strong education system, preferably with a media literacy component. We also need news organisations that maintain high standards of evidence and fact-based reporting.

On the other hand, we would like to see more fact-checking on social media in the most automated way possible. Our experience has shown us that fully automated fact-checking is not yet feasible. I am not even sure it is advisable because, despite all the clamour around AI, I still think it fails in terms of accuracy in fundamental aspects. We see this in the hallucinations and fabricated sources that AI provides us with to generate reports and studies – which is nowhere near the level of human intelligence in evaluating high-quality evidence. Finally, fact-checkers are not against technology – many of them use AI in their own work. That is why I say that if AI is used by humans, it must also be supervised by humans. It cannot simply be left free online.
 

HQ: How can ordinary people cope with so many conspiracy theories and unverified news stories, especially when the sources are as many as they are now?

AH: I think we are struggling. We live in an oversaturated online information ecosystem, and there is no peaceful way out of this. If society works together collectively, we can improve the situation. But at the moment, there seems to be very low levels of social trust. I do not think the technology platforms are doing as much as they can to try to make online information better for the average user. They are just prioritising their corporate goals of maximising profit on a large scale. There are no magic solutions here. I believe we have decades of work ahead of us, trying to keep the knowledge systems that humanity has developed over hundreds of years in a good state and bring them into the online space. There are some positive points, though, such as university libraries that have done a fantastic job of making information and archives available online. Research communities have also shared studies and findings online across international borders. Also, Europe is taking a strong regulatory approach, and we want to see what they will do with the implementation of the Digital Services Act. It is not all a bad story.
 

HQ: In other words, we have more high-quality information sources than ever before, but the channels and platforms that distribute this information are not necessarily aligned with the common good. Is that what you mean?

AH: Yes, but we also have to think about the question of what expertise is. For specialists, there has never been so much information available as there is now, which leads them to do more in-depth research. But for non-experts, the environment is very challenging. And, let's be clear, nobody is an expert in everything. So, whatever your area of expertise - mine is fact-checking - there are elements beyond our remit. Whenever you enter an area in which you have no expertise, you need to be very cautious. That is more or less what we are seeing now.
 

HQ: Do you think these social media platforms should be regulated at government level?

AH: I believe they need to be held accountable to the public. In many countries, this will depend on the context of the regulation. In others, it may be more through legal action or the legal system. I do not want to present myself as a public policy expert, especially given the variety of legal and social contexts that so many countries in the world have. I would adopt a moderate position, which says that they need to be held accountable. If certain countries and contexts believe that regulation is the way to do it, that should not be ruled out. In our global community, the Europeans are trying to develop a very strong regulatory framework, while the Americans, who tend to have lighter regulation, focus a lot on lawsuits. In Brazil, on the other hand, the judicial systems are taking a very tough stance against disinformation.

What I am saying is that there is no single way of dealing with the problems of disinformation, fraud, rumours and conspiracy theories. There are several ways of dealing with it. I believe we need to give all these experiments a chance to be realised in different countries and see how they work. Concerns about freedom of expression are legitimate, as we have seen authoritarian countries try to use laws banning disinformation against their political opponents. It cannot be said that these concerns are unfounded. Sometimes they are expressed very cynically and without any real intention of improving the information space, but they cannot be dismissed out of hand.
 

HQ: Do you think the traditional media are falling behind on these issues by trying to outpace social media in their own lane?

AH: It is a mixed picture. We can see that many global networks have maintained their standards and the quality of their news screening. However, we see plenty of other media outlets that have gone completely partisan and have abandoned the ethical standards they used to have. There are also these new phenomena that often look like journalism, but without any fact-based information values, such as online influencers. These new digital actors are extremely idiosyncratic. Some of them spread rumours and conspiracy theories; others actually broadcast high-quality information; and others seem to be very heterogeneous. There is nothing at first glance that allows the casual user to evaluate and make a decision so quickly. You have to delve into the content, analyse it carefully and do your own research. And that's what makes online information experiences so exhausting for so many people.
 

HQ: How can an AI collaborative approach increase productivity and save time, without tarnishing the ethics and transparency of fact-checkers?

AH: I think AI agents are very good at synthesising languages. If you need to produce a report based on content you are familiar with, generative AI tools are very good at putting it into different formats and compressing large amounts of data. Now, you cannot rely on AI for factual accuracy because it often produces misleading outcomes. We hear stories in the press all the time about professionals relying on AI to generate raw content, and it always backfires. Humans need to police AI to ensure the quality of data and information. You can take a long article and put it into an AI assistant, to break down and create all kinds of content and findings for social platforms. Some professionals are experimenting with it in different formats, turning stories into podcasts for example. Once again, when the author is sufficiently familiar with the content, these are tools that streamline and refine the quality of our various tasks. There was a time when we thought that if we ignored bad information, it would disappear, but this is no longer the case.
 

“What seems to be happening is that algorithms are outperforming high-quality content in their ability to capture people's attention”
 

HQ: Angie, your IFCN post highlights a new fact-checking initiative responding to a surge of online disinformation about several African countries called TruthAfrica. Could you briefly explain the project and what it has revealed about the gap between African and European audiences in their perception of independent journalism?

AH: Just to clarify that this is a project by two members of our network, so I cannot take any credit for it. What happened was that the fact-checkers in Poland began to realise that a lot of the disinformation they were seeing was about Africa. Disinformation narratives about Africa were being disseminated to the European public. This was part of a political agenda aimed at creating division between Europe and Africa, or suggesting that Africa was wasting development money. Every disinformation narrative has its own little incentive behind it. Polish fact-checkers then teamed up with their African colleagus to create new reports and debunk what was emerging first-hand.

In this sense, the core of the project was to inform the European public in an accurate and verified way about what was really going on in Africa. This is the kind of collaboration that would not be practical without the online space. The fact-checkers talk to each other via email and text messages, and share the text reports electronically so that they can edit them. These are the advantages that technology brings. I thought it was an admirable project as it utilised these cross-border collaborations to get to the bottom of what is verifiable and what is not.
 

HQ: Do you have a final word for our association community? How can they protect themselves from dubious and false information?

AH: I would say that it's very important to speak out against false information and correct it with correct information. I think there was a time when we thought that if we ignored bad information, it would just go away, but that has not been our experience lately. So now I'm encouraging everyone to be very proactive and speak up when they see something that is not right. The other thing I would say is that we all need to work together in the knowledge-seeking community. We all need to support each other, regardless of what area we work in. I think the support of colleagues at this time is very important.

Published by Meeting Media Company, the publisher of Headquarters Magazine (HQ) – a leading international publication based in Brussels, serving the global MICE industry and association community.