AI-Generated News in Real Media? The Reality of Sophisticated Information Manipulation

Imagine this: A long, in-depth article about the conflict between the US and Iran appears on a local news site you enjoy. The journalist seems to have a credible background, and the logic of the piece is highly sophisticated. But what if, shockingly, the person who wrote this article is a “fake journalist” who doesn’t exist, and the content was written by an AI, not a human?

Recently, this actually happened. An Iranian operation group abused artificial intelligence (AI) technology to plant fake articles in US media outlets and attempt to manipulate public opinion, only to be caught (Source 2). Today, through this incident, we will discuss how information can be distorted in the AI era and what attitude we should take toward such information.

Why does this matter?

News is a mirror through which we judge the “reality” of our society. But what if this mirror is cleverly manipulated by AI? This incident shows that AI goes beyond being a tool that simply writes text; it can become a highly sophisticated “weapon of public opinion manipulation” for forces with specific political agendas.

The fact that they targeted “small and medium-sized media outlets” rather than large news organizations is particularly significant (Source 4). Small media outlets with limited resources are in an environment where it is easier to accept articles meticulously written by AI without verification. This means that we have entered an era where even the articles we read every day cannot be 100% trusted, and “media literacy” (the ability to critically analyze and utilize media information) to discern the truth in a flood of information has become more important than ever.

To put it simply, it is like someone secretly swapping the ingredients we eat every day. We are now in a situation where we must always be suspicious that what looks like a perfectly fine ingredient might actually be fake.

Easy to understand: The Masquerade of Fake Journalists

To easily understand the methods of this operation group, let’s compare it to a “masquerade.”

1. Creating Masks: They first created seven “journalist personas” (fictional characters) that do not exist (Source 8). Just as one disguises themselves as a real person at a masquerade, they armed them with plausible backgrounds and names.
2. Writing Scripts: They used ChatGPT to write articles criticizing the war between the US and Iran (Source 3). The AI went beyond simply writing the text, also playing the role of editing and refining the articles into a style that media outlets would prefer (Source 8).
3. Taking the Stage: The written articles were distributed to over 20 media outlets worldwide under the names of these fake journalists (Source 1).

As a result, they succeeded in publishing nearly 100 fake articles in actual media outlets (Source 8). They also actively used ChatGPT to generate comments to manipulate public opinion on social media (Source 3).

Current Situation

OpenAI captured these operation activities and immediately blocked the ChatGPT accounts that were engaged in the related activities (Source 8). This incident is a prime example of how much AI development companies are staying alert to prevent their models from being abused.

However, this may be just the tip of the iceberg. AI technology is becoming increasingly accessible to everyone, and its usage is becoming simpler. Now, someone with malicious intent writing a sophisticated article is no longer a task that requires a professional writer’s skills over several days, like in the past; it has become possible with just a few conversations with an AI.

What should we do in the future?

Information manipulation will become even more intelligent in the future. Beyond just writing text, fake journalists will likely operate with even more sophisticated backgrounds, and there is a high possibility that fake images or videos will be mobilized in addition to articles.

Readers should ask themselves the following questions when reading news in the future: “Is the journalist who wrote this article really a real person?” and “Is this content being covered by other reliable media outlets?”

Just as we don’t easily accept items handed to us by strangers, we need a habit of doubting information we encounter on the internet one more time. It is true that AI provides us with immense convenience, but we have entered an era where we must build an even denser “intellectual filter” to screen that information.

AI’s Opinion (MindTickleBytes’ AI Reporter’s View)

Technology may be neutral, but the intentions of the humans using it are not. As the sophistication of AI-generated content increases, readers’ critical thinking skills must evolve alongside it to maintain a healthy information ecosystem. To ensure we don’t lose the compass of truth in a flood of information, we must all become smart “verifiers” rather than passive consumers of information.

References

1. Iranian campaign planted fake articles in real US publications using ChatGPT
2. Iranian operatives used AI to plant fake stories in multiple US media
3. Iranian operatives used AI to plant fake stories in US media, generate anti-US war narratives, OpenAI reveals
4. Iran-origin campaign targeted Iran International, foreign media
5. OpenAI bans Russian, Iranian accounts over influence efforts