2026 rw deepfakes rsf
Digital safety

Dealing with Deepfakes (1/3): practical recommendations for journalists

This post is part 1 of 1 in the series Deepfakes

Propelled by the spectacular rise of Generative Artificial Intelligence (GAI), deepfakes — digital manipulations that impersonate real people — are now commonplace in  the global news landscape. Between December 2023 and December 2025, Reporters Without Borders (RSF) identified at least 100 journalists targeted by this technology across 27 countries with the documented cases representing only the visible part of a much larger phenomenon. What three forms can this threat take? And, how can they be detected?

Using a Familiar Face to Scam People

The most common method of creating a deepfake involves using people’s trust in a well-known media personality to manipulate them . There are two common scenarios

In the first scenario, the journalist’s image is used without their consent to “sell” a product: cosmetics, dietary supplements, financial services, etc. In the second, the journalist is placed in a staged situation, such as a fake arrest or altercation, in order to encourage viewers to click on a fraudulent link.

This type of attack is widespread and is often perpetrated using  sponsored posts. A sole journalist may be bombarded with  hundreds of fake pieces of content simultaneously. Victims are rarely warned by a pre-existing alert system: instead, they are contacted by relatives, unsuspecting  viewers, or angry strangers when they realise they have been scammed.

Television and digital media journalists, whose faces are familiar to the general public, are very often the target of such attacks. 

Manipulating Public Opinion

Deepfakes designed to spread disinformation exploit journalists’ images in a more deceptive manner The aim is to make false information appear credible or to discredit a journalist by attributing false statements to them

In Portugal, the face and voice of presenter Pedro Benevides were manipulated in a video published on Facebook in September 2025. The fake video made him claim that the government had conspired with the pharmaceutical industry during the Covid-19 pandemic.

Another example took place in September 2023, when an AI-generated audio clip simulated a conversation between Slovak journalist Monika Todova and a political leader, giving the impression that they were about to rig Slovakia’s parliamentary election. The deepfake went viral despite a media blackout on the eve of the election.

Harassment of Journalists

The deepfakes used to harass journalists create other challenges as their goal is not  to deceive, but to humiliate. The content serves as ammunition for groups that are already hostile towards the journalist — a spark that can ignite a smear campaign.

In the last few  years journalists have been portrayed as jihadists or pigs, or wearing military uniforms in volatile wartime contexts. This category also includes pornographic deepfakes: women journalists depicted in entirely fictitious sexual situations.

These deepfakes are designed to intimidate and silence journalists by making the personal cost of doing their job unbearable.

To go further