With the introduction of the Internet, the opportunity to access information from around the world grew exponentially. At the same time, however, it also provided a vaster forum for misinformation and falsehoods. Social media has helped expand its reach in this regard, creating echo-chambers of exaggerations and mistruths. But these look to be simply steppingstones to the future. Already, AI-generated fake news abounds on the Internet and even in print media. And disinformation campaigns using AI are on the rise. As these tools become even more advanced, it’s going to be increasing hard to tell what’s fact and what’s fiction. This is particularly concerning when information remains a major currency in the world today.
(Generative AI has been fed a steady diet of Internet nonsense–read more in this Bold story.)
Notably, there are many motivations for using AI-generated fake news. Some of these motivations are political in nature, especially when major elections have the potential to change global power structures. Others are economic in nature as disinformation campaigns using AI can enhance opportunities for advertising dollars and viewership. Regardless of the incentive, however, the end result is not a promising one. Erosion of trust in media is already widespread, and AI-generated fake news will only make things worse. It’s therefore worth exploring what’s likely to lie ahead when it comes to the future of media overall.
The Growth of AI-Generated Fake News
It’s hard to believe generative AI platforms were only introduced to the public a little over a year ago. Their impact is already being felt in numerous industries, and it’s expected usher in major changes in the immediate future. Unfortunately, however, one of these changes relates to AI deepfakes and AI-generated propaganda. According to one misinformation watchdog, Newsguard, nearly 750 AI-generated news sites now exist. Each of these tend to lack any oversight to ensure accuracy of information. And some simply exist to promote disinformation campaigns using AI for malicious reasons. Unless scrutiny over these sites and their reporting is used, there is the potential for some serious social harms.
(AI-generated deepfakes are a problem–read up on it in this Bold story.)
One example of AI-generated fake news recently involved a report involving the Israeli Prime Minister. The story alleged that the Prime Minister’s psychiatrist had committed suicide to draw attention to the insanity of current Israeli policies. But while this may have reflected a satirical perspective, many news sites reported it as truth. Some even went as far to use AI-generated fake news to create the psychiatrist’s obituary. In investigating the story, its similarity to a satirical piece dating back to 2010 in a Pakistani paper was noted. This explained how it could have been a disinformation campaign using AI that was trained on that article. This is just one example of thousands that now exist. And it’s going to become only more commonplace over time as fake news sites expand.
The Good Getting Better
Over the last year, a number of companies, artists, and content generators have leveraged AI in their work. What seemed to start out a little rough has certainly improved in a short amount of time. When it comes to written content, generative AI is now being used to create full-length news pieces by major media outlets. Buzzfeed reported earlier this year that it was going “all-in” on AI to assist with its media content. Sports Illustrated was even found to have used fake AI-generated writers who had fake headshots as well. As AI improves in its ability to compose higher quality content, this shows it will be harder to identify falsehoods. This is especially true for those who quickly scan headlines and the first few sentences of a story. The power that disinformation campaigns using AI have could therefore be significant.
One of the current concerns in this regard is not related to written content but to video images. Though not yet available to the public, the video generator Sora could make a powerful impact regarding AI-generated fake news. It has the ability to make extremely high-quality videos that are Hollywood quality. As such, it becomes increasingly hard to tell what has been AI-generated and what is reality. Disinformation campaigns using AI tools like Sora could show completely false live footage to support their perspectives. This adds a new dimension to AI-generated fake news that makes it that much more difficult to discern fantasy from reality. If a picture paints a thousand words, then AI-generated fake footage’s impact could be tremendous.
Free Speech Versus Fake Propaganda
When it comes to AI-generated fake news, sites involved tend to hide behind the protection of free speech. Imposing legal restrictions to prevent disinformation campaigns using AI constantly runs into this issue. Even social media companies cannot be held responsible for misinformation on their platforms based on this doctrine. At the same time, those involved in these campaigns are difficult to actually identify, leveraging anonymity behind digital technologies. And even if they are, most are not within national borders. As a result, policing AI-generated fake news is not something that’s likely to occur at a governmental or even international level. Instead, it’s going to be the public as well as media sites themselves that will provide the oversight needed.
In a year ahead with major elections worldwide, it can be expected that numerous disinformation campaigns using AI will appear. Russia is already creating AI-generated fake news as it relates to Ukraine and its leaders. And some U.S. legislators are reciting this information as they discuss the issues. This demonstrates that power that AI might have over public opinion in time. While using AI to generate news stories in itself isn’t a problem, fake news is. It will therefore be essential to access such stories from reputable media that police content ethically. And it will also be important to verify such reports by sampling a number of other sources. We now live in an era where AI-generated fake news will be rather prevalent. And the only way to best determine truth and reality will require an increasing level of scrutiny and discernment.
Programming an AI to Have Biases Creates a Whole Slew of Problems–Read Why in this Bold Story!
