As the business landscape evolves rapidly, enterprises are accelerating the integration of generative AI (GenAI) into their operations to enhance efficiency and competitiveness. However, without proper attention to cybersecurity risks, GenAI could become a new gateway for hackers. According to a Gartner report, by 2027, over 40% of AI-related data breaches will be caused by improper use of GenAI across borders.
With growing awareness of AI governance, many vendors have launched AI security solutions, yet most focus only on model accuracy or hallucination detection, overlooking the cybersecurity vulnerabilities of the models themselves. Leveraging its deep expertise in AI risk research, AIFT launched Vulcan, a security platform built for GenAI. The platform provides an end-to-end solution, from assessment to protection, helping enterprises strike a balance between innovation and security.
AIFT's unique AI model risk assessment technology has already completed proof-of-concept (POC) trials with more than five major financial institutions. In 2025, AIFT was publicly recognized by the Singapore government during Personal Data Protection Week for its innovative achievement in multilingual GenAI risk identification, underscoring its international credibility and technological strength.
Leveraging AWS cloud technology to build a hybrid cybersecurity solution
With extensive technical experience in cybersecurity, AIFT was among the first to identify the emerging risks associated with GenAI. In response, the company created Vulcan, an end-to-end platform that empowers enterprises to manage AI innovation and cybersecurity simultaneously.
The Vulcan platform features two core modules: Vulcan Attack and Vulcan Protect. The former simulates red team attacks, enabling enterprises to identify vulnerabilities in GenAI systems from a hacker's perspective and perform risk analysis and remediation before deployment. The latter continuously monitors AI model inputs and outputs, detecting behaviors that violate internal policies or compliance requirements, thereby preventing sensitive information leakage.
By integrating these two modules, enterprises can ensure both security and compliance in AI deployments. With AI scenario-based attack techniques, Vulcan Attack shortens the traditional 200-hour risk assessment process to just 3 hours, significantly boosting deployment efficiency. Vulcan Protect, meanwhile, offers high flexibility for customized testing and can be seamlessly integrated into GenAI environments and AI governance frameworks, ensuring uninterrupted operations.
According to Frank Liao, Director of Partnership at AIFT, since joining the Amazon Web Services (AWS) Joint Innovation Center (JIC) in 2023, AIFT has received strong technical support from AWS experts, which has significantly optimized Vulcan’s computing performance and resource management. This year, Vulcan Attack has been officially launched on AWS Marketplace, allowing global enterprises to conduct POC validation and testing in hybrid architectures. With AWS JIC's support, AIFT has expanded its global reach and connected with a broader base of potential customers.
Expanding footprint in Asia-Pacific and the Middle East with Multilingual GenAI Security Assessment Capabilities
As GenAI applications evolve from single-language to multilingual environments, AIFT’s cross-language risk identification capability has become a key competitive advantage in the international arena. For highly sensitive sectors such as finance and high-tech manufacturing, AIFT can tailor configurations and security rules according to each enterprise’s internal policies and compliance standards, ensuring alignment with regulatory and industry requirements.
Frank emphasized that AIFT is one of the few companies capable of assessing AI models in multiple languages. Beyond English and Chinese, AIFT supports Japanese, Korean, and Thai for Asian markets, as well as Arabic and Kazakh for Middle Eastern customers. With its multilingual detection expertise, AIFT plans to deepen its presence across the Asia-Pacific region and further expand into the Middle East, leveraging its successful cases in Taiwan and AWS resources to help more clients complete complex security assessments and simulation POC projects at lower cost.