Published: Apr 24, 2025
enhancing phishing validation through AI-driven solutions
Phishing scams have become one of the most pressing cybersecurity challenges, with losses in Singapore reaching a staggering $1.1 billion in 2024[1]. As phishing tactics continue to evolve, traditional detection methods are proving insufficient. These outdated approaches often fail to accurately assess whether a website is legitimate, leading to delayed responses and increased risk. Recognising this urgent need, the NUS-NCS Lab developed an advanced phishing validation solution that uses deep learning and AI-driven multi-layered content analysis to determine the legitimacy of websites with greater precision and efficiency.
The challenge:
Phishing attacks have become increasingly sophisticated, making it harder to accurately validate a website’s authenticity. The rise of encrypted web traffic and the use of generative AI allow phishing websites to bypass conventional detection. Many no longer rely on obvious tactics like copying logos, instead using subtle and deceptive branding that slips past standard filters. This growing complexity demands a smarter, faster validation approach—one that goes beyond traditional blacklists and manual checks. The core challenge is ensuring high validation accuracy while reducing manual workload and response times.
The solution:
The solution is built on a combination of visual annotation and object-label detection, enabling it to assess whether a website genuinely belongs to the organisation it claims to represent. By analysing domain name consistency, targeted brand phishing intent, and content structure, the system significantly improves validation precision. Unlike conventional models that rely heavily on human intervention, this approach automates the validation process, reducing human effort by 85% and enabling the rapid processing of up to 100,000 suspicious URLs daily. This latest advancement incorporates generative AI to automate validation, enhance detection accuracy, and strengthen security. It also ensures that phishing scams are flagged more swiftly, preventing potential financial and data losses.
Snapshot of capabilities:
- Multi-layered content analysis combining image and text checks to validate websites
- Use of generative AI to validate website authenticity
The impact:
By developing cutting-edge phishing validation technologies, the NUS-NCS Lab, in collaboration with the NCS cybersecurity team, is leading the way in combating online scams. The solution has led to significant improvements in validation precision, dramatically reduced manual effort, and increased the speed of verifying suspicious websites. This innovation strengthens digital trust for businesses and individuals alike, ensuring they remain protected in a fast-moving threat landscape. It marks a major step forward in proactive cybersecurity and sets a new benchmark for website validation and resilience against phishing.
[1] Source: The Straits Times