Forecast data from cybersecurity experts at Kaspersky state that in 2026, the Asia-Pacific (APAC) region will become a global leader and significant testing ground for artificial intelligence (AI) innovations. It was found that AI usage among experts will reach as high as 78%, surpassing the global average of 72%. The main drivers are younger consumers and widespread access to numerous technological devices, leading AI innovations to rapidly permeate daily life and business sectors.
However, this progress is a double-edged sword that is reshaping the direction of cyber threats. AI will shift from being a supplementary tool to becoming the core component of both attacks and defenses in full.
The most concerning trend in 2026 is Deepfake technology becoming a primary tool for fraud, as barriers to accessing the technology decrease, allowing even non-experts to create highly realistic synthetic images and audio with just a few clicks. The development of real-time online Deepfakes will become more seamless, enabling manipulation of virtual camera videos to deceive in live communications. Amid challenges with content labeling systems lacking reliable standards and being easily fooled or removed—especially in open models—this will blur the line between legitimate advertising content and deceptive material, making it difficult to distinguish by the naked eye or common detection systems.
In the cyberattack chain, AI will be used in nearly every phase—from preparation, vulnerability discovery, malware coding, to generating highly sophisticated phishing websites and fraudulent emails. Large language models (LLMs) will assistthreat actorsto automate tasks more extensively and better conceal traces of their attacks.
Furthermore, the rapid growth of open-source models lacking strict controls comparable to closed models will become a key channel for threat actors to customize and create highly effective attack tools with wide dissemination.
Conversely, AI technology will play an essential role in transforming security operations centers by transitioning to agent-based systems that can continuously scan infrastructure and analyze threat contexts 24/7. This will reduce routine workload and enable experts to make more accurate decisions through natural language interfaces, replacing complex technical commands with everyday language interactions, thereby accelerating detection and mitigation of complex threat incidents.
To cope with the evolving security landscape, organizations must proactively adapt through key measures such as regularly updating software to close vulnerabilities, restricting remote desktop accessfrom public networks,and backing up data separately from main networks.
Additionally, raising awareness and training employees to understand the risks of Deepfake and synthetic media will become a crucial security agenda that every organization must systematically implement to harness AI innovation while maintaining sustainable cybersecurity.