SilverFox, one of the most active advanced persistent threat groups in the Asia-Pacific region, is exploiting the rapid adoption of artificial intelligence in businesses, with Kaspersky warning that AI is helping transform cyberattacks from operations requiring specialized teams into low-cost campaigns that can potentially be launched by individuals.

Kaspersky Global Research and Analysis Team researchers first detected the threat actor in December 2025, identifying a multistage payload delivery system supported by segmented infrastructure that uses different addresses and domains throughout the attack chain.

The approach is designed to reduce the chances of detection and prevent security defenses from blocking an entire campaign after identifying one malicious component.

SilverFox most recently targeted companies in India, Indonesia, South Africa and Russia across the industrial, consulting, trade and transportation sectors, using phishing emails disguised as official tax audit notifications or messages urging recipients to download an archive supposedly containing a list of tax violations.

The campaign exploited the authority and urgency associated with tax agencies to persuade recipients to open the malicious files, with Kaspersky recording more than 1,600 malicious emails between January and February 2026.

More recently, Kaspersky GReAT researchers found that SilverFox had begun distributing fake Claude applications as another way to infiltrate target organizations.

Claude, developed by Anthropic, is an advanced conversational assistant and large language model that can help users write, code, analyze lengthy documents and solve complex problems through natural-language interactions.

“SilverFox is one of the most active threat groups in the whole APAC region. They get into targets through three simple routes: fake websites, phishing emails, and harmful files spread via social messaging apps. They inject malware used for long-term cyberespionage and sensitive data gathering. Our recent analysis showed they are now distributing fake Claude for Windows, macOS, and Linux, leveraging AI use in companies to crack into their targets’ security defenses,” explains Ye Jin (Seth), lead security researcher at Kaspersky GReAT.

The Asia-Pacific region accounts for the largest share of SilverFox’s malicious activity by a wide margin, exceeding the combined volume recorded across all other regions and indicating that the group’s current operations are heavily concentrated in Asia, particularly East and Southeast Asia.

“Based on our current threat data, Greater China is SilverFox’s main target with over 90% of all its attacks targeting the region. Mainland China alone makes up 71%. Myanmar, Cambodia and Singapore also see lots of attacks. These are the next hotspots we need to watch,” adds Ye Jin.

Manufacturing is SilverFox’s most targeted industry, accounting for more than one-third of attacks, while IT and services follow closely amid extensive phishing activity aimed at technology workers.

Healthcare and finance are also considered significant targets because of the value of the information and systems handled by organizations in those sectors.

AI changes the speed of cyberattacks

Kaspersky’s analysis also highlighted JADEPUFFER, described as the world’s first fully large language model-driven ransomware, as an example of how AI can move beyond assisting attackers to making decisions and carrying out attacks itself.

As the first agentic ransomware, JADEPUFFER demonstrated how AI can analyze situations, adapt tactics and execute attacks in real time rather than relying entirely on human operators to determine the next step.

“In this particular case, disclosed by our Sysdig, the malicious AI agent completed the entire cycle of diagnosing a failed attempt, correcting its approach, and launching a new attack in just 31 seconds, far outpacing the response capabilities of most human defenders. Beyond speed, JADEPUFFER also demonstrates an unprecedented level of autonomy. It shows that AI agents are now capable of making independent decisions throughout the attack process, effectively replicating the reasoning of an experienced human attacker without direct oversight,” he explains.

AI-driven attacks are also creating new challenges around stealth, particularly through techniques that exploit users’ trust in established AI services.

Ye Jin cited ChatGPhish, an indirect prompt injection technique that hides malicious instructions inside webpages and uses trusted AI web-summarization functions to relay those instructions when users ask an AI assistant to summarize the content.

Because the malicious instructions are delivered through a trusted AI interface, victims may be more likely to follow harmful links or instructions, while conventional security tools can have difficulty identifying the activity because it resembles legitimate interaction between a user and an AI service.

AI is also lowering the technical barriers to developing sophisticated malware, according to Kaspersky.

The discovery of VoidLink, an AI and cloud-native malware framework, in January 2026 demonstrated how generative AI can assist in the development of complex malicious software, with the framework reportedly created almost entirely with AI assistance.

Kaspersky calls for AI-native defenses

Kaspersky said organizations need to respond to AI-powered threats by adopting defensive technologies and strategies capable of matching the speed and adaptability of modern attacks.

The company identified four approaches for organizations seeking to strengthen their defenses:

• Proactive defense: Use AI-driven threat hunting to identify unknown threats rather than relying solely on passive responses.
• Zero Trust architecture: Strictly verify every access request to reduce risks associated with implicit trust inside networks.
• Systematic defense: Build comprehensive protection across endpoints, networks, applications and data.
• AI vs. AI: Use large models and dual-use AI technologies to strengthen detection and response while continuously adapting to attackers.

“When attackers can leverage AI to automate decision-making and accelerate every stage of an attack, defenders must respond with the same level of intelligence. Cybersecurity solutions enriched with continuously updated threat intelligence are no longer a competitive advantage, but a critical requirement for staying ahead of rapidly evolving threats,” adds Ye Jin.

Kaspersky is a global cybersecurity and digital privacy company founded in 1997 that uses a Cyber Immunity approach to protect consumers, businesses, critical infrastructure and governments from cyberthreats.

The company said its technologies have protected more than 1 billion devices to date and that it focuses on protecting revenue, reducing workloads and preventing downtime through cybersecurity solutions and services for organizations of different sizes.

Kaspersky said it is trusted by millions of individuals and nearly 200,000 organizations and recognized in independent tests, while its security technologies combine AI-driven protection, threat intelligence, simplified management and expert support.