Key Takeaways

  • Data Theorem added AI-driven capabilities that form a continuous loop from exploit discovery to runtime defense
  • Market forecasts project the application security testing sector to reach $5.1 billion in 2025 as AI-generated threats multiply
  • The platform targets cloud-native, API-heavy environments where source code is often incomplete or unavailable

Data Theorem unveiled a new set of AI capabilities that bring its application and API security portfolio into a closed-loop model. The announcement focuses on AI Exploits, AI Auto-Remediation, and AI Active Protection, each managing a distinct stage in discovering, fixing, and defending against AI-generated attack chains. Enterprises are reconsidering how quickly risks propagate across modern applications, as security teams increasingly encounter attacks assembled by large language models.

The launch aligns with rapid expansion in the application security testing market. According to the Gartner Magic Quadrant for Application Security Testing, the sector is projected to reach $5.1 billion in 2025. This growth tracks a fundamental shift in software architecture: enterprise applications increasingly run on AI, introducing risks that traditional security tools were not designed to handle.

The AI Exploits capability addresses discoverability by running against a live application rather than the source tree. Production workloads rarely map cleanly back to a single repository or build artifact, as multiple microservices or libraries typically feed into a deployed service. Reverse engineering and dynamic inspection can surface paths that static reviews overlook. The platform's Analyzer Engine blends static, dynamic, and binary analysis along with model reasoning to simulate attack chains rooted in real behavior instead of theoretical code paths.

While frontier models have made source-code exploit discovery more accessible, production applications rarely arrive with perfectly reconstructable sources. AI Exploits closes this gap by working against the production-grade application as it runs, removing the dependency on supplying perfect source code to reconstruct reachable exploit chains.

AI Auto-Remediation addresses the challenge of long patch cycles. Detection creates noise unless risks are actively reduced, and attackers can now use AI to chain exploits faster than engineering teams can patch them. This feature prioritizes urgent vulnerabilities, generates recommended fixes, and pushes changes into production pipelines automatically. While organizations often insert manual review checkpoints for critical logic or sensitive components, moving from a detected exploit chain to a completed patch with minimal delay targets the rapid emergence of AI-generated vulnerabilities.

On the runtime side, AI Active Protection leverages the company's established API Protect and Mobile Protect SDKs. Because patching alone cannot always keep pace with discovery, runtime controls provide organizations space to operate while fixes are pending. AI Active Protection focuses on guardrails, behavior analysis, and detection patterns for issues such as scraping, memory probing, prompt injection, and data exfiltration. Implementing runtime enforcement requires balancing active defense with the need to prevent false positives from disrupting applications.

The broader market includes multiple vendors investing heavily in cloud-native security and API protection. This platform differentiates by extending beyond AI-powered discovery into a continuous loop encompassing automated remediation and active runtime defense. This sequence offers an integrated approach for environments where source code is fragmented.

Enterprise security teams often require extensive testing before standardizing on workflow automation, particularly when AI participates in code modification. However, the presence of established runtime SDKs currently in production use provides a tested foundation for organizations evaluating these automated capabilities.

AI influences both sides of the security equation, with attackers utilizing chained exploit generation and defenders responding with systems that neutralize risks in near real-time. Data Theorem's release aligns with this shift. The integration of closed-loop platforms across application security depends on maturity, trust, and performance under real conditions. These capabilities target organizations seeking to shrink their exposure windows and keep pace with automated threats.