Lessons for brands from the frontlines of social media manipulation

July 6, 2023

Written by NewsWhip

Governments globally, particularly in conflict zones, confront the persistent threat of social media manipulation, as evidenced by ongoing information warfare in regions like Ukraine. However, there are dedicated organizations such as Valent Projects actively working to detect and counter disinformation by malicious actors on a global scale.

At NewsWhip’s June Linkedin Live event, Amil Khan, CEO and founder of Valent Projects, joined NewsWhip’s CEO, Paul Quigley, to address effective strategies for combating digital manipulation and highlight the critical role of real-time monitoring.

Amil Khan, a former journalist, and his team at Valent Projects identify and track digital manipulation to prevent real-world harm. They map out networks and infrastructure used by different actors in conflicts, leveraging tools like NewsWhip to obtain valuable insights into ongoing events. 

“We use tools such as NewsWhip Spike to predict how content will get shared to get a heads up on what is happening,” Khan explains.

By monitoring the velocity of posts on public pages such as Meta, Valent Projects uncovers relationships and connections, giving them a comprehensive understanding of the digital information landscape. This helps their clients to take proactive action in responding to emerging news and combat manipulation effectively.

State sponsored social media manipulation

Valent Projects has played a crucial role in supporting Meta to uncover instances of state-sponsored media manipulation, as they did in 2019 when they identified posts that indicated a civil war was imminent in Sudan.

Khan explains, “We have, in the past, been able to identify these kinds of networks at a high enough evidence level for the Meta team to take it and do their own checks and decide to take action.”

Despite the active presence of Meta, the rapid pace of AI content creation still poses significant challenges. As AI becomes more complex, Valent Projects has been developing AI detection software to address this evolving landscape, enabling them to distinguish between machine-generated and human-generated text.

How AI-generated content impacts misinformation

The ability to generate different versions of content using AI tools presents a challenge. Khan emphasizes the need to have tools capable of consistently detecting generative AI across various types of content. For instance, the emergence of AI-powered tools like Adobe’s text and image generators, while impressive, raises concerns about manipulation. 

“Its text-generative, image-generative AI looks absolutely amazing, but very scary from a manipulation point of view,” Khan said.

The sophistication of AI-generated content has reached a level where it closely resembles human-created material, making it resource-intensive to identify manipulation.

Information overload from AI

Lastly, Quigley brings up the issue of information overload and noise generated by AI, emphasizing the challenges it poses in discerning reliable content from the abundance of available information. In response, Khan expressed that AI-generated content may appear impressive, but in this context, you want to think of this as a premium product versus a factory-generated product. 

For instance, he speculated that if AI were capable of producing novels, there would still be a missing human factor that provides extra depth and a unique perspective. Khan remarked, “If novels were produced by AI at scale, you can bet someone will sit down and write really good novels which will be better and have that extra element that can’t be generated by a machine.”

Khan also emphasizes that if people encounter an overwhelming amount of AI-generated content, they will disengage, as “when there is so much junk, people will just switch off.” He firmly believes that the public’s preference would naturally gravitate towards reliable sources providing quality reporting.

This highlights the ongoing importance of quality journalism in an era where AI-generated content can overwhelm and confuse audiences.

If you’d like to learn more about Valent Projects and their disinformation-fighting tool, read our blog here.

You might also like

Six key features for newsrooms seeking a CrowdTangle replacement

Six key features for newsrooms seeking a CrowdTangle replacement

With CrowdTangle being deprecated in August 2024, many newsrooms are seeking an alternative for their social media news gathering, multichannel audience development, competitive benchmarking and developing ideas from social media trends. I recently spoke to Chris...


Related articles

The NewsWhip Note

Regular updates from the NewsWhip team on the crises, trends, and stories that are resonating with the media and public around the web.