‍

As “AI slop” - low-quality, high-volume digital content generated by artificial intelligence designed to distract, seduce, and deceive - increasingly dominates our feeds, newsrooms are under way more pressure to deliver not just the news; but also the truth.

‍

So we sat down with Sophie Nicholson, Head of Innovation & AI Projects, AI Literacy at Agence France-Presse (AFP), to hear her approach to navigating the age of careless content and how AFP are doubling down on human-centric verification to overcome the daily battle with synthetic media.

‍

Accuracy over speed is becoming the new standard

‍

The traditional journalistic drive to be first is being re-evaluated in the face of synthetic media. Sophie explained that for a global agency like AFP, being right is the only option.

‍

"... verification is really at the heart of all our journalism. And It's a key guideline to all our teams to verify … even if you're not going to put something out as quickly as you would like to".

‍

This represents a shift from traditional journalism when being first mattered - whether legacy broadcast, newspapers, or 24/7 websites. Organizations like AFP are stepping back a little bit due to the implications and risks of being wrong. 

‍

No "magic bullet" - but human judgement becomes more critical

‍

While there is a high demand for tools that can automatically detect AI-generated content, Nicholson warned that no such perfect solution exists.

‍

"The thing that we tried to be really clear about when we're communicating on AI is … there is no silver bullet. There's no tool that's going to verify 100% AI generated content, even though we're starting to see tools that can help".

‍

Instead of relying solely on software, AFP encourages journalists to return to "traditional" reporting. Nicholson noted: "It's all about focusing on seeing things with your eyes and being on the ground and just going back to that kind of traditional way of getting information".

‍

And this validates data we saw first hand in our “Future newsroom: How GenAI is reshaping journalistic skills” report which revealed that, while new technical skills are becoming increasingly necessary, so-called “traditional” competencies like source verification, fact checking, and critical thinking are becoming more important than ever.

A practical checklist for combating AI slop

‍

One of the many key takeaways from the talk was Sophie sharing AFP’s "do’s and don’ts" that govern their journalists’ application of AI:
‍

• Always verify everything: This includes links, facts, and sources - verify, verify, verify!
  ‍
• Experiment in low-risk areas: Play with and explore AI tools, but do not do it with your live work.
  ‍
• Be transparent: Maintain transparency about the process and any experimentation.
  ‍
• Don’t anthropomorphize AI: Remember that it is a tool, not a source; do not quote it as one.
  ‍
• Protect sensitive data: Do not share confidential data with AI chatbots, as they are designed to keep you "adding information".
  ‍
• Don’t be lazy: Avoid the risk of becoming lazy, such as letting AI write first drafts
  ‍

A note on transparency

‍

As the internet faces what some call "enshitification" - the decay of platforms into "slop platforms full of ads and contentless content", the value of credible, face-to-face journalism increases. Sophie believes that transparency, even when mistakes happen, is the key to longevity.

‍

"It's impossible not to make mistakes but just to be very transparent about them... the fact that you're having a conversation and you're being transparent about what you're doing and how you're trying to do it... you're authentic".

‍

Catch up with our conversation around AI slop by listening back to the Trint transcript here, or chat to our team if you’re interested in how Trint can help your organization verify live content.