Introduction
The 2025 API ThreatStats Report by Wallarm highlights a dramatic increase in AI-driven API vulnerabilities, emphasizing the critical role of APIs as a primary attack surface due to the widespread adoption of AI technologies. The report underscores the significant security risks posed by API flaws, which are linked to nearly all AI-related vulnerabilities.
Description
AI-driven API vulnerabilities have surged by 1,205% over the past year [2] [3], as highlighted in the 2025 API ThreatStats Report by Wallarm. This report underscores that APIs have become the primary attack surface [1] [4] [5] [7], largely due to the rise of AI adoption [6], which significantly contributes to security risks [5] [7]. Notably, nearly 99% of AI-related vulnerabilities are linked to API flaws [2] [3]. Key findings reveal that 57% of AI-powered APIs are externally accessible [1] [3] [5] [6] [7], while 89% utilize insecure authentication methods [1] [2] [5] [6] [7], leaving many endpoints exposed [1] [5]. Alarmingly, only 11% of these APIs have implemented robust security measures [2] [3], exposing organizations to severe risks [7], including technical vulnerabilities and potential reputational damage [7].
In 2024 [2] [3] [7], Wallarm identified 439 AI-related Common Vulnerabilities and Exposures (CVEs) [3] [5] [6] [7], marking a staggering 1,025% rise from the previous year [1] [4] [5]. These vulnerabilities primarily stem from injection flaws, misconfigurations [1] [2] [3] [4] [5] [6] [7], and a newly recognized category of Memory Corruption and Overflow [3], which addresses issues arising from improper memory handling in AI workloads [6]. This category is particularly concerning due to AI’s reliance on high-performance binary APIs, which can lead to unauthorized data access [1] [4] [5], crashes [4], and arbitrary code execution [4]. The increasing deployment of AI technologies has been a significant driver of these vulnerabilities, with 53% of enterprises engaged in multiple AI projects [2].
Furthermore, for the first time [4] [7], over 50% of recorded vulnerabilities exploited by CISA were API-related [5] [6] [7], reflecting a 30% increase from the prior year [4] [5] [6] [7]. This trend emphasizes the critical importance of API security [4] [5] [7], as vulnerabilities in this area now surpass traditional exploit categories such as kernel [5], browser [1] [4] [5] [7], and supply chain vulnerabilities [4] [5]. Both legacy and modern APIs are under attack; legacy APIs remain vulnerable due to outdated designs [7], while modern RESTful APIs face risks from complex integrations and misconfigurations [7]. Notable exploit types include improper authentication [7], injection attacks [6] [7], and endpoint misconfigurations [1] [6] [7], with legacy APIs accounting for over 18% of exploited vulnerabilities [7]. Additionally, over 33% of CISA’s Known Exploited Vulnerabilities (KEVs) involve modern API technologies.
The report cites specific breaches [7], such as those involving Twilio and Tech in Asia [7], illustrating how weak authentication and access control mechanisms can be exploited [7]. The frequency of API-related breaches has tripled in 2024, escalating from a few incidents per quarter in 2023 to multiple occurrences each month [2]. Overall, the findings stress that AI security is intrinsically linked to API security [4] [7], necessitating real-time API controls to protect business operations and customer trust as organizations increasingly rely on APIs for innovation and transformation [7]. Prioritizing API security is essential for organizations to safeguard their systems and fully leverage APIs as a catalyst for business transformation [4].
Conclusion
The findings of the 2025 API ThreatStats Report underscore the urgent need for organizations to prioritize API security as a fundamental component of their cybersecurity strategies. The rapid increase in AI-driven API vulnerabilities poses significant risks, including technical vulnerabilities and reputational damage [7]. To mitigate these risks [2], organizations must implement robust security measures [7], such as real-time API controls and secure authentication methods. As APIs continue to drive innovation and transformation, ensuring their security will be crucial for protecting business operations and maintaining customer trust in the future.
References
[1] https://www.prnewswire.com/news-releases/wallarm-releases-2025-api-threatstats-report-revealing-that-apis-are-the-predominant-attack-surface-302357760.html
[2] https://osintcorp.net/ai-surge-drives-record-1205-increase-in-api-vulnerabilities/
[3] https://www.infosecurity-magazine.com/news/ai-surge-record-1205-increase-api/
[4] https://itnerd.blog/2025/01/29/wallarm-releases-2025-api-threatstats-report-revealing-that-apis-are-the-predominant-attack-surface/
[5] https://ai-techpark.com/wallarm-releases-2025-api-threatstats-report/
[6] https://channelvisionmag.com/wallarm-apis-are-the-predominant-attack-surface/
[7] https://cioinfluence.com/security/wallarm-releases-2025-api-threatstats-report-revealing-that-apis-are-the-predominant-attack-surface/
