By Patrick Coffee

Bot networks are now a brand problem, too.

Coordinated webs of fake social-media personas have historically been the domain of fraudsters and state-based actors, such as the infamous Russian troll farms that spammed influential personalities like X owner Elon Musk, because of the time and labor involved.

But the rise of generative artificial-intelligence tools has made botnet campaigns both easier to manage and harder to detect.

Coordinated attacks targeting corporations have grown much more common over the past year or two, with bot networks going after brands such as Amazon, Target, McDonald's and Boeing, said Emilio Ferrara, a professor of computer science and communication at the University of Southern California who has studied social-media bots for years.

"What's different now is how quickly AI-powered bots can spin up 'grassroots-looking' campaigns around incendiary or divisive issues, e.g., culture-war topics, and keep them trending," Ferrara said.

"Attribution is hard, but these examples are illustrative of nonstate campaigns directed at brand reputation," he added.

Ascertaining motivations for these attacks can be as difficult as identifying the individuals behind them, but money is a top contender, according to PeakMetrics, a startup that tracks online discussions and threats for clients including government agencies and companies.

Verified users on X that meet minimums for followers and impressions can get paid based on the likes, replies and shares their posts get from other verified users. Bots masquerading as humans can help meet those minimums and spread posts so they reach more people who might engage, PeakMetrics said.

Botnets zero in on divisive topics, all the better to trigger human reactions, and aren't limited to any particular platform or political ideology. And they can make complaints loom larger than they would on their own, putting executives on the defensive.

Cracker Barrel Old Country Store recently abandoned a new logo under criticism from consumers, influencers and elected officials including President Trump. The episode was only the restaurant chain's latest brush with customers upset over its turnaround strategy, and Chief Executive Julie Felss Masino acknowledged on the company's latest earnings call that the brand had "misstepped."

But a disproportionate share of the social-media chatter that fed the news cycle about Cracker Barrel's logo was driven by bots on X, according to PeakMetrics, which works with the U.S. Air Force to identify sources of foreign misinformation, as well as Cyabra, an Israeli disinformation-detection firm.

The online logo backlash began with human-run accounts that have hundreds of thousands of followers. They started encouraging a boycott Aug. 20, one day after Cracker Barrel first mentioned its new logo in a press release, often accusing Cracker Barrel of fleeing its country charm and past.

Hundreds of bots soon began sharing the posts, replying to them and posting their own, PeakMetrics said.

The height of activity came just before midnight on Aug. 20, when X saw around 400 Cracker Barrel posts a minute. Seventy percent of the accounts promoting boycotts at that point used duplicate messages, a key marker of coordinated bots, said Molly Dwyer, director of insights at PeakMetrics.

Other signs included repeating the same thing dozens of times or posting consistently almost 24 hours a day, Dwyer said.

Bots or likely bots authored 44.5% of X posts mentioning Cracker Barrel in the 24 hours after the new logo gained attention on Aug. 20, according to research for The Wall Street Journal by PeakMetrics. That number rose to 49% among posts calling for a boycott.

The share of posts that typically come from bots and likely bots during discussions of controversial issues on X runs from around 20% to 30%, PeakMetrics said.

X's parent, xAI, didn't respond to requests for comment. Cracker Barrel declined to comment.

Prominent political accounts, such as that of Sen. Marsha Blackburn's gubernatorial campaign, eventually shared their own takes on the controversy. Trump used his Truth Social to weigh in against the new logo, and on Aug. 26, Cracker Barrel reversed course.

Meanwhile, recent bot network attacks have magnified posts advocating boycotts of Amazon and McDonald's because the companies rolled back diversity, equity and inclusion programs, according to Cyabra, the disinformation-detection firm.

Cyabra's clients have included the U.S. State Department, entertainment giant Warner Bros. Discovery and public relations agency Golin. Elon Musk hired the firm to gauge bot activity on the former Twitter before he bought the social platform in 2022.

Cyabra checks 600 to 800 parameters including location, posting frequency and the use of AI-generated avatars to declare whether accounts are human or not, said Chief Marketing Officer Rafi Mendelsohn.

Tracking bots doesn't necessarily help stop attacks. "If you're the brand, if you're the CMO of the brand, what the hell do you do?" Mendelsohn said.

Still, knowing that these negative attacks are coordinated by third parties could help marketers avoid engaging directly with the bots or mistaking every complaint for a human's.

Executives should consider any decision they make a potential tool for activity designed to heighten negative engagement on social media, said Dwyer, the Peak Metrics executive.

"There is an ecosystem of people who are looking for anything to grasp onto," she said. "There has been a vibe shift, and I don't think that companies have quite caught up to it yet."

Write to Patrick Coffee at patrick.coffee@wsj.com

(END) Dow Jones Newswires

September 26, 2025 06:00 ET (10:00 GMT)

Copyright (c) 2025 Dow Jones & Company, Inc.