Jailbroken AI behind Russian cyber campaign against American politicians

Jailbroken AI Russian cyber campaign was used in an advanced operation targeting US politicians and politically engaged individuals, according to new information from TrendAI. The campaign allegedly included phishing, fraud, and AI-generated influence operations, using large language models to create content, automate workflows, and scale cyberthreats on a larger scale.

New data from TrendAI shows that jailbroken large language models, known as LLMs, have been used to create advanced cyberthreats targeting US politicians and politically engaged individuals. The campaign reportedly included AI-generated influence operations, credential theft, phishing, and large-scale cryptocurrency fraud.

According to the research, it involves a sophisticated cyber campaign called fire campro, which is said to have been run by a lone Russian-speaking threat actor. The campaign is said to TrendAI have been built using a jailbroken version of Google Gemini and other AI-based tools to automate content creation, set up phishing infrastructure, collect login credentials, and run cryptocurrency scams.

AI was used to scale up the cyber campaign

TrendAI's analysis shows that AI technology was not only used to generate content, but also to streamline the entire cybercriminal infrastructure behind the operation.

The campaign is said to have used AI, among other things, to:

• create compelling political content and advocacy messages
• automate phishing campaigns and fake login pages
• process stolen API keys and automated workflows
• conduct crypto fraud on a larger scale with lower operational costs

According to the report, the cyber campaign has built an audience of approximately 17,000 followers on Telegram, while AI-powered automation tools and stolen API keys were reportedly used to greatly reduce the costs of the attacks.

A new type of AI-driven cybercrime

TrendAI believes that brandcampro illustrates how modern AI systems can be used to create a new generation of scalable cyber threats where influence, automation and financial crime are combined in the same operation.

TrendAI believes that this jailbroken AI Russian cyber campaign shows how AI can be used to scale up cybercrime in a way that previously required significantly greater resources.

– This cyber campaign shows how modern AI systems enable a new type of scalable and cost-effective cybercrime that combines influence campaigns, automation and financial fraud, says Martin Fribrock, Country Manager Sweden, Finland and Baltics at TrendAI.

Jailbroken AI models create new risks

The case also highlights growing concerns about so-called jailbroken AI models, where security restrictions in large language models are circumvented to create content or tools that would otherwise be blocked.

When such models are used by threat actors, AI can not only generate text, but also help build phishing pages, automate attacks, create fake identities, and run advanced fraud at significantly lower costs than before.

Security experts warn that jailbroken language models could create a new type of cyberthreat where attackers combine disinformation, phishing, financial fraud and digital influence in a single operation.

New risk picture for AI security

Security experts warn that jailbroken language models could create a new type of cyberthreat where attackers automate phishing, fraud, disinformation, and digital influence on a much larger scale than before.

When AI is used to create content, build fake login pages, and drive automated workflows, cybercriminals can lower their costs while making attacks more sophisticated and harder to detect.

For companies, governments and political organizations, this development means that AI security is now becoming an increasingly important part of cybersecurity work. It is not only about protecting data, but also about understanding how generative AI can be used as an offensive tool by threat actors.

At the same time, the development shows that AI security is no longer just a technical issue for IT departments. It is increasingly becoming a strategic issue for management, risk management and national security.

Read more about TrendAI's research on fire prevention here: