Asset Publisher

Single title

Reporting on AI-Driven Disinformation, A Working Guide for Journalists

New Study : When Fakes Are Cheap, the Spread Is the Story

This study by Dr. Heather A Thuynsa, Yanga Malotana and Dr. Christopher Nehring is an instructive and effective guide for journalists chasing deadlines to help identify, verify and report manipulated content without amplifying it.

Asset Publisher

Generative AI has made fabricated video, audio and text easy to produce. But more convincing fakes are only part of the challenge. AI has also made disinformation cheap enough to produce and distribute at scale. For journalists, the key question is therefore no longer only whether something is fake, but who made it, who pushed it and how it spread.

The new guide by Heather A. Thuynsma, Yanga Malotana and Christopher Nehring, Reporting on AI-driven disinformation: A working guide for journalists, offers reporters and editors a practical method for identifying, verifying and reporting manipulated content without amplifying it. Developed with Project Vote at the University of Pretoria, it was initiated and funded by the KAS Media Programme Sub-Saharan Africa.

 

A Guide Built for the Deadline

Designed for use under newsroom pressure, the guide is organised around eight practical questions, checklists, decision tables and four quick-reference cards. Its starting point is simple: Is the content false? Was it meant to deceive? Who made it, and who pushed it?

A seven-step verification workflow takes journalists from preserving suspicious content to documenting what can and cannot be established. Importantly, the guide cautions against over-reliance on AI detection tools, particularly for compressed, re-uploaded and non-English material. Provenance, context and human expertise remain essential.

The guide also asks a crucial editorial question: Should the false claim be reported at all? When coordinated operations flood the information space with multiple fabrications, journalists should focus on the campaign and its distribution networks rather than chasing every individual claim.

 

Beyond Verification

Corrections alone are not enough. They need a distribution strategy capable of reaching the audiences exposed to the original falsehood. The guide therefore looks at WhatsApp and other closed messaging networks, local languages, trusted community messengers and the discoverability of corrections through search engines and AI assistants.

It also addresses information-disorder burnout and the responsibility of newsrooms to provide journalists with training, verification time, legal support and recognition of the psychological burden of this work.

Drawing largely on African elections and information environments, the guide reflects the realities of mobile-first news consumption, closed messaging and limited platform moderation in local languages. Its central message is straightforward: interrogating the claim is no longer enough. Increasingly, the spread is the story.

 

 

 

Asset Publisher

Contact

Rebecca Sibanda

Rebecca Sibanda
Project Manager
rebecca.sibanda@kas.de +27 (11) 214 2900 +27 11 214 2913/4

comment-portlet

Asset Publisher