OpenAI shuts down ChatGPT accounts used by organized scam networks

OpenAI has shut down a network of ChatGPT accounts linked to a sophisticated criminal operation believed to be running scams from Southeast Asia.

Key takeaways

• OpenAI disrupted ChatGPT accounts linked to an organized criminal network that allegedly used AI to support romance scams, investment fraud, gambling scams and law enforcement impersonation schemes.
• The group used ChatGPT to create fake personas, translate conversations, generate promotional content and assist with internal operations, helping scammers scale their activities.
• OpenAI's investigation also uncovered indicators consistent with human trafficking and forced criminality inside some scam compounds.
• The findings reinforce trends highlighted in Bitdefender's Global Scam Intelligence Report 2026, which found that modern scam operations increasingly function like organized businesses that rely on automation and specialized roles.

How scammers used ChatGPT

Generative AI is a valuable productivity tool that helps millions of people across the globe write emails, summarize documents and translate text in seconds.

Unfortunately, cybercriminals have adopted the same tools, using AI to defraud millions of people.

According to OpenAI, the disrupted network relied on ChatGPT to support several fraud schemes at once, including romance scams, fake investment opportunities, gambling scams, and law enforcement impersonation.

“Earlier this year, we disrupted a Cambodia-based scam operation that used ChatGPT to support investment, romance, gambling, and law enforcement impersonation schemes,” the report reads. “We began investigating this activity following a lead from our peers at WhatsApp and have since shared additional threat signals with industry partners and relevant authorities.”

Rather than carrying out the scams automatically, AI helped operators work faster and reach more potential victims.

The report says the group used ChatGPT to:

• Create convincing fake online identities
• Generate and translate conversations into multiple languages
• Write promotional content and advertisements
• Draft investment and romance messages
• Prepare internal documentation and operating procedures
• Help with administrative tasks

The findings provide a rare glimpse into how organized cybercriminals are integrating generative AI into their day-to-day operations.

You may also want to read:

• Lost money to a scam? FBI warns about recovery scams
• Global Scam Crackdown: 276 Arrested, Crypto Fraud Networks Dismantled

Scam operations are highly organized

One of the most striking aspects of OpenAI's investigation is that it wasn't dealing with a handful of individual scammers.

The same organization allegedly operated multiple fraud schemes simultaneously, sharing infrastructure, personnel and technology across different scams.

That's consistent with findings from Bitdefender's Global Scam Intelligence Report 2026, which describes today's scam operations as increasingly sophisticated organizations with specialized roles, management structures and scalable workflows rather than loosely connected criminal groups. AI has simply become another tool that helps these operations work more efficiently.

“For instance, operators used dating personas to build trust before introducing fraudulent investment opportunities involving cryptocurrencies and spot gold trading,” Open AI said. “Other users engaged in lengthy romantic conversations with targets using fictitious identities, posed as representatives of online gambling platforms offering fake bonuses and winnings, or impersonated law enforcement agencies to tell targets they needed to pay fines for committing serious criminal offenses.”

Beyond the technology, OpenAI's report also highlights a darker reality.

The company identified conversations containing references to fake overseas recruitment, debt bondage, salary deductions, restricted movement and attempted escapes. While OpenAI says it can’t independently verify each case, these conversations are consistent with years of reporting on scam compounds operating in parts of Southeast Asia.

Bitdefender's Global Scam Intelligence Report also noted large-scale enforcement actions targeting scam compounds in countries such as Cambodia and Myanmar, where authorities have dismantled criminal operations while also identifying workers believed to have been trafficked or coerced into participating in online fraud.

The problem extends beyond Southeast Asia. South African authorities recently warned job seekers to be wary of fake overseas employment offers after traffickers were found targeting people with promises of legitimate jobs, only to force some victims into criminal scam operations abroad.

AI may change the tools, but not the tactics

Although AI can make scams more convincing, the techniques used to manipulate victims remain largely the same.

Scammers still rely on building trust, creating urgency and convincing victims to send money or share sensitive information.

Whether the scam begins with a dating app, a social media message, an investment opportunity or someone claiming to represent law enforcement, the warning signs are often familiar:

• Unexpected contact from strangers
• Promises of guaranteed profits or unusually high-paying jobs
• Pressure to move conversations to private messaging apps
• Requests to act quickly
• Demands for cryptocurrency, gift cards or wire transfers
• Reluctance to let you verify claims independently

How can consumers stay safe?

Before responding to unexpected messages or investment opportunities:

• Verify the person's identity through official channels
• Research companies and recruiters independently
• Avoid clicking unsolicited links
• Be skeptical of offers that seem too good to be true
• Never send money because someone creates a sense of urgency

Layered security can also help stop scams before they succeed. Features such as Scam Protection, Scam Radar, Email Protection, SMS Protection, Web Scam Protection, Remote Access Scam Protection, and free AI-powered tools like Scamio and Link Checker can help identify malicious links, phishing websites and scam messages before they succeed.

You can find all these and more in our all-in-one security solutions.