One in 6 financial market accounts is fake, study finds

Fake accounts that inflate or deflate specific stocks and amplify boycott sentiment against financial institutions are spreading across global financial markets — roughly one in every six accounts. Similar risks may be emerging in South Korea, where more than 20 percent of accounts participating in online discussions about financial fraud have been detected as fake, analysts say.

According to global intelligence firm Cyabra, the share of fake accounts in online discussions about global stock markets rose from 13 percent in 2024 to 17 percent in 2025 — meaning roughly one in six accounts is classified as fake. Over the same period, engagement generated by these accounts increased 50.8 percent and their potential reach grew 52.6 percent.

Fake accounts moved to either promote or attack specific stocks. Cyabra's analysis identified 1,188 fake profiles promoting Warren Buffett's investment firm Berkshire Hathaway. In the opposite direction, a single fake account posted more than 17,400 messages repeatedly attacking Tesla — Elon Musk's electric vehicle company — targeting its sales figures, technology and Musk's leadership as CEO.

Cases of fake accounts amplifying negative sentiment against financial institutions were also found. An analysis of the backlash that spread after JPMorgan flagged MicroStrategy's bitcoin exposure risk found that 23 percent of participating accounts were fake. The potential reach of related content was estimated at about 3 million people.

The fake accounts Cyabra refers to are not simply automated "bots." The term covers any account displaying inauthentic characteristics in identity or behavior — including AI-generated personas, accounts using false identities or impersonating others, and accounts operating under multiple identities.

The concern is that when multiple fake accounts repeat the same claim or react to each other's posts, they can make it appear that far more people hold a given opinion than actually do. "A thousand complaints from real customers and a thousand negative posts created by a small number of actors using hundreds of accounts carry entirely different meanings, but online they can look like public opinion of a similar scale," Cyabra said. The firm sees the difficulty of distinguishing genuine user sentiment from artificially amplified voices as a key risk factor in financial markets.

Fake account activity has also been detected in South Korea. Cyabra analyzed online discussions on X related to financial fraud — covering topics such as personal data leaks, card fraud, loans, investment advisory chat rooms and virtual assets — from Aug. 18 to Wednesday. Of 709 profiles examined, 145, or 20.45 percent, were detected as fake.

Among the actual posts, many emphasized high returns — phrases such as "made 2 million won ($1,450) in profit in two months" and "looking for someone to earn over 2,000 a month" appeared frequently. Others were designed to lure users to external channels or investment advisory chat rooms, with messages like "the password for the chat room is 666" or "send a message saying 666."

Fake account participation was also detected in conversations related to the forgery and sale of identity documents, bank books and passports. Among the key behavioral traits identified in fake accounts, automated behavior — bots — was the most common, accounting for 40.72 percent.

However, the fake account activity detected in South Korea has not been confirmed to involve the kind of organized financial market manipulation seen overseas. "We identified some connections between accounts and similar behavior patterns, but found no clear evidence of a single campaign in which one actor systematically amplified a specific financial issue," Cyabra said.