We talked to Diana Leaños, a fact-checker for El Sabueso at the Mexican media outlet Animal Político, about how she works and what important aspects she wants you to know about fact-checking.
What are your first steps of checking a source for facts?
One of the first things to consider in order to fact check something is to look into the context in which our claim is being said. Whether we are fact checking a viral claim online or a politician, one needs to look into the context. Who is saying it and why? And what is being said, is it possible? Then I check various media outlets, non-governmental organizations, government information, or whatever the case may require. This could even include interviews with specialists.
Do you know any fact-checking websites that can help checking a source?
Yes. When I have to fact check something that is non Mexico related and there is no information in major international media I check the International Fact-Checking Network (IFCN) signatories. This unit created by the Pointer Institute thoroughly selects the agencies and media outlets that conform it and any organization that is part of the network follows a specific methodology that is committed to fact checking.
Social Media is currently flooded with AI, how can you be certain if a video of news is AI or not? What’s the easiest way to find that out?
AI still has many giveaways. The first thing to notice are visible watermarks from AI chatbots like Grok and Sora. If these things aren’t visible we can move on to inconsistencies in the video, especially if we are talking about one that portrays humans. AI still can’t produce hands accurately or even replicate normal breathing and blinking. In this case this should be the first giveaway. Next AI can’t focus on many things at once, look into the background and you may be able to notice shapes that blend into one another or things that magically appear. Another good hint are short videos, AI is still very expensive and can only reproduce short videos that last 10 to 15 seconds maximum.
When these things aren’t clear you can use an AI detection tool. Google has SynthID, which can detect the hidden watermarks that are left by Google’s AI, Gemini. Another good tool identifier is HiveModeration. Still when using these tools you should preferably check with more than one analyzer, because sometimes they can fail and give you a false positive.
We want to create a checklist for students so they can go along and check their source themselves, what aspect can not be missing on it?
Invite people to ask themselves:
☐ Where does this information come from?
Knowing where the supposed information came from is very important to determine whether it is a reliable source or if it is trying to deceive us.
☐ Who says it and what evidence do they have?
If the supposed news does not include a date, place, the names of the people involved, the full names of the authors of the alleged scientific study that supports the information, or the source or authority that produced the data, it is likely to be misinformation.
☐ What do they gain if I believe it?
People who spread misinformation seek to benefit from deceiving us, so it is important to understand what would happen if we believe them or why they need us to believe them. This could be financial gain through “clickbait,” electoral gain, or monetary profit from selling “miracle products” that actually harm our health.
☐ How does this make me feel?
Misinformation feeds on our emotions, especially strong ones. If I see something that perfectly matches what I think, makes me feel so happy or angry that I need others to know immediately, or makes me want to act to protect myself from something that could be misinformation.
☐ Is anyone else talking about this?
If it seems true but you are still unsure, search on Google or your usual browser to check the same information elsewhere. You might find missing details, such as the source or explanation of the facts, in other media outlets or specialized websites. On the other hand, if no other news outlet, authority, or official page is sharing that information, it is likely to be misinformation.
by: Carmen Schmidt