Challenges from AI
AI (artificial intelligence) continues to explode on an unprecedented scale. By the end of 2025, ChatGPT had reached 810 million monthly users (MAU), while Gemini also reached nearly 750 million people. This number is equivalent to 10% of the world's population. According to Deloitte, 99% of organizations are forced to change their operating models to apply AI. When AI becomes a universal infrastructure, it will be both a motor and a cyber attack front.
Talking to Lao Dong Newspaper, Ms. Nguyen Quynh Anh - Marketing Director of VinCSS Cyber Security Company emphasized: "AI is becoming the optimal tool for hackers. 2026 will be a year of further boom for hackers to use AI to automate the entire life cycle of attacks: from collecting information, exploiting vulnerabilities, creating malware, jailbreaking, producing deepfaked, fraudulent content with low cost, high speed and greater scope of influence than ever before.
In parallel with that, cybercrime is increasingly being "industrialized" through many different models, even those without technical background can become hackers.
Meanwhile, Mr. Ngo Tan Vu Khanh, Country Director of Kaspersky in Vietnam, predicts that the cybersecurity picture in Vietnam in 2026 will become increasingly complex, especially under the impact of AI.
Mr. Vu Khanh said: "Three threats are not only stopping at fraud or data theft, but are shifting strongly to cyber espionage and targeted sabotage, with AI used to automate attacks, personalize scripts, support "agentic" malware to operate autonomously and adapt in real time, avoiding traditional detection mechanisms.
According to Mr. Dang Huu Son, Co-founder & CEO of LovinBot AI, besides many opportunities, AI brings many cybersecurity and digital fraud risks. “Scam forms using AI, deepfake videos, voice and image imitation are becoming much more sophisticated as new generation video creation models such as Google's Veo 3 or Kling Motion are becoming increasingly popular.
This not only affects individuals, but also directly threatens the reputation and assets of businesses if they lack appropriate knowledge and defense systems," Mr. Huu Son shared.
Every citizen is a cybersecurity warrior
Faced with increasing cybersecurity challenges from AI, according to Mr. Ngo Tan Vu Khanh, Vietnam needs to approach cybersecurity comprehensively: from raising user awareness, increasing investment and business response capacity, to protecting key infrastructure and developing specialized cybersecurity human resources.
Mr. Khanh also highly appreciated the birth of the Hanoi Convention on Cybercrime Prevention. He said: "The Convention creates a common legal framework for the prevention and combat of cybercrime, especially cross-border cases, thereby helping functional agencies and Vietnamese businesses have a clearer basis to coordinate investigation, share information and respond to incidents".
According to Ms. Nguyen Quynh Anh, the Convention will only be truly effective if it comes with technological capacity, implementation capacity and human resources. “We have heard many mentions about the issue of protecting key infrastructure, national data, and business data.
However, how to protect, what to protect is still an open problem. Personally, I am particularly interested in three directions: post-quantum cryptography to protect long-term data; safe access management according to Zero Trust, Passwordless/Passkey/FIDO; and AI application to "fight back" AI," Ms. Quynh Anh shared.
Ms. Quynh Anh also raised the issue that cybersecurity cannot only rely on experts. Vietnam is still lacking about 700,000 cybersecurity experts. “That forces us to change our mindset: each digital citizen needs to become a minimally "cybersecurity warrior.
Building a network security culture, democratizing network security and making it a responsibility of each individual, that is the most sustainable defense for the AI era," Ms. Quynh Anh emphasized.
