March 2026 Summaries

Southeast Asia is rapidly advancing in digital transformation by integrating AI and cloud-native architectures, which has shifted focus from merely adopting the cloud to managing complex software supply chains. This evolving landscape, characterized by the proliferation of compiled binaries, containers, and machine learning models, presents challenges in visibility and governance. JFrog's partnership with iZeno Pte Ltd aims to address these issues by providing tailored solutions that meet the diverse regulatory and operational requirements of the region. This partnership focuses on ensuring compliance and security across distributed environments, reducing tool sprawl to enhance developer efficiency, and integrating rigorous management of AI models within the software supply chain. By doing so, it empowers organizations in Southeast Asia to innovate with confidence, transforming their software supply chains from vulnerabilities into competitive advantages.

The LiteLLM supply chain compromise of March 2026 highlights an evolving threat landscape where attackers shift focus from developers to AI agents that developers rely on, showcasing a new level of sophistication in supply chain attacks. This incident, orchestrated by TeamPCP, involved compromising the LiteLLM package, a critical component in the AI ecosystem, by exploiting unpinned dependencies in its CI/CD pipeline to inject malicious payloads. These payloads harvested sensitive credentials and attempted lateral movement across systems, demonstrating the significant risk associated with AI infrastructure packages. The attack underscores the vulnerabilities inherent in open-source AI gateways, which serve as central routing layers between applications and LLM APIs, making them attractive targets for attackers. In response, JFrog emphasizes the importance of governing AI agentic supply chains with enterprise-grade controls, introducing solutions like the JFrog AI Gateway and MCP Registry to enforce security policies and prevent such compromises. These tools aim to replace blind trust in open-source components with a secure, policy-enforced framework that ensures AI assets are vetted before integration, aligning with the broader vision of a Trusted Agentic Supply Chain.

Security policies are critical for protecting software supply chains, yet they often disrupt builds, causing frustration among DevOps and security teams. This challenge arises when security measures, meant to prevent supply chain attacks, block packages due to policy violations, leading to failed builds and wasted developer time. The solution is not to relax policies but to enhance tooling, as demonstrated by JFrog Curation's Compliant Version Selection (CVS). This capability finds and serves the highest policy-compliant package versions automatically, transforming security enforcement from a roadblock into a seamless process. Unlike traditional binary approaches that halt builds, CVS offers a block-and-serve model, ensuring that developers receive compliant packages without workflow interruptions. This not only maintains security without impeding development but also saves significant developer time, enhancing trust in security policies. CVS is integrated into JFrog Curation, applying instantly across ecosystems, and extends its governance to AI/ML models, ensuring robust policy compliance without additional burden on developers.

JFrog is repositioning itself as a critical player in the evolving landscape of software supply chains, which are increasingly driven by AI and require robust security and governance. At a recent event in NYC, JFrog introduced nine innovations aimed at securing and managing AI agents, ensuring enterprise scalability, and maintaining global resilience. These innovations span two main areas: Security & Governance for the Agentic Frontier and Enterprise Scale & Global Resilience. They include tools like the MCP Registry for secure agent connections, the Agent Skills Registry for managing agent capabilities, and Policy-as-Code for flexible governance. Additionally, JFrog has enhanced capabilities for repository federation and smart retention to streamline operations across global teams, along with introducing native support for the Nix Package Manager and offering a premium 99.99% uptime SLA. These developments underscore JFrog's commitment to being a trust layer in the AI-driven software delivery landscape by integrating governance, security, and compliance into the core infrastructure where binaries reside.

In 2026, the rapid pace of AI-driven software development has introduced a new risk visibility gap, as organizations struggle to manage the security and compliance challenges of AI-generated and third-party code snippets. Developers frequently incorporate code from AI prompts and open-source components, leading to potential security vulnerabilities and legal compliance issues, such as viral license risks and hidden vulnerabilities that evade traditional software composition analysis (SCA) tools. To address these challenges, JFrog has introduced Code Snippet Security, a feature enhancing JFrog Xray's capabilities by using semantic matching to detect risky code fragments. This innovation offers enterprises enhanced security and compliance by identifying hidden vulnerabilities and restrictive licenses, ensuring software integrity and maintaining a verifiable audit trail. By integrating this capability into development workflows, organizations can mitigate risks while leveraging AI-generated code, thereby turning potential liabilities into strategic advantages.

In a world increasingly reliant on AI and the Model Context Protocol (MCP) for connectivity, the security and governance of AI systems are paramount, as a breach in MCP servers can lead to severe risks for enterprises. The JFrog MCP Registry addresses these concerns by providing a unified control plane for managing and securing MCP servers within the Agentic Software Supply Chain. This registry functions as the sole source of truth, allowing enterprises to enforce rigorous security measures such as granular access controls and automated policy gates, thereby blocking malicious servers and unauthorized commands. By treating MCP servers as managed software artifacts, the JFrog MCP Registry enables seamless integration with AI-native IDEs and ensures that only compliant servers are executed, reducing the risk of supply chain attacks. This approach allows organizations to adopt AI technologies without compromising security, effectively balancing innovation with enterprise-grade security and compliance standards.

As the AI agent ecosystem evolves, skills have emerged as a crucial component for equipping agents with domain-specific expertise and automating workflows. Skills are lightweight, reusable units of knowledge that can be easily adopted to transform agents from generalists into specialists by embedding organizational know-how into their operations. However, the proliferation of skills introduces challenges similar to those faced with open-source software, such as security risks and governance issues. JFrog addresses these concerns by offering the Agent Skills Registry, which provides centralized management, enhanced security, and comprehensive control over skills, enabling enterprises to scale AI systems without compromising safety or compliance. This registry ensures that skills are treated as first-class software assets, offering features like version control, provenance generation, and secure consumption, which are essential for preventing the emergence of unmanaged dependencies and security vulnerabilities.

Azure Machine Learning (AzureML) offers robust model experimentation and compute capabilities, but many organizations face challenges in transitioning models from development to production securely. The process is often hindered by unmanaged silos, which can lead to issues like lack of traceability, security vulnerabilities from unvetted packages, and compliance gaps. To address these challenges, the integration of AzureML with the JFrog Software Supply Chain Platform is proposed, creating a governed AI pipeline that treats AI assets as standard software artifacts. This integration involves a four-step workflow, ensuring that every model and dependency is securely scanned, versioned, and managed through a unified supply chain, thus bridging the gap between AI development and enterprise execution. By employing this approach, organizations can maintain the security and governance of AI models while facilitating their path to production, ensuring compliance and reducing risks associated with Shadow AI.

As AI continues to transform the software development lifecycle, it brings both increased productivity and significant security challenges, according to a discussion between JFrog's Jens Eckels and Forrester's Janet Worthington. They noted the rapid evolution of application security (AppSec), emphasizing the necessity of consolidating security tools due to tool sprawl and vulnerability backlogs. The rise of AI in software development has created a new landscape where AI-generated code introduces potential vulnerabilities that require rigorous governance and quality assurance to prevent security breaches, particularly in the software supply chain. Forrester data reveals that external attacks, including software supply chain breaches, significantly impact organizations, resulting in high costs and productivity losses. The discussion highlighted a shift in the market towards holistic security platforms, like the JFrog Software Supply Chain Platform, which aim to streamline security processes, reduce vulnerabilities, and increase productivity by providing a unified interface and context for developers. By implementing solutions such as JFrog Curation, organizations can preemptively block insecure packages, fostering a secure development environment and reducing the burden of human code reviews.

In an era where development teams are innovating rapidly with the help of generative AI coding assistants and modular architectures, the reliance on third-party components like open source packages and AI models has increased, leading to potential vulnerabilities in software supply chains. Attacks such as the Shai-Hulud and React2Shell highlight the risks associated with these dependencies, as they exploit gaps in security measures. As organizations face the challenge of balancing speed and security, there's a shift towards integrating security measures at the point of entry for third-party components, using automated, policy-driven controls to evaluate and vet dependencies before use. This approach aims to manage risks without hindering development workflows, especially in AI-driven environments, by ensuring that only compliant and secure components are used, ultimately enabling both velocity and governance in modern software development.

In today's global enterprise environments, the outdated approach of centralized computing is giving way to more distributed, federated systems, especially in software development. Many organizations still rely on legacy infrastructure for managing their binaries, resulting in issues like slow downloads, version drift, and high data transfer costs. JFrog addresses these challenges with its intelligent Repository Federation, which offers solutions such as Active-Active synchronization for seamless global collaboration, high availability for disaster recovery, and granular control for compliance and cost management. By bridging on-premise and cloud infrastructures, JFrog Federation ensures developers can access data with local-speed efficiency, eliminating the latency and productivity losses associated with remote repositories. This approach not only enhances operational resilience but also aligns infrastructure with modern distributed workflows, turning previously fragmented systems into a cohesive, high-speed network.

In the realm of modern software development, a stark contrast exists between the well-governed software supply chain and the chaotic AI supply chain, often leading to security vulnerabilities and governance issues. While traditional software development follows a secure and automated process with CI/CD pipelines and rigorous scanning, AI development frequently operates in a disconnected manner, with developers using public hubs and local machines, creating a "shadow supply chain." This blog post, the first of a five-part series, addresses the need for consolidation as the initial step towards trusted AI adoption. By merging AI and software assets into a single, governed lifecycle, companies can eliminate off-road practices, ensuring AI models and components are treated with the same security and governance as traditional software. This consolidation involves using centralized proxies for model downloads, creating an AI registry, and linking AI components to enterprise software for full traceability, ultimately transforming governance from a hindrance into a competitive advantage. The post sets the stage for further exploration of managing hidden AI connections in future installments of the series.

Recent incidents highlight Continuous Integration (CI) workflows as prime targets for software supply chain attacks, with vulnerabilities in GitHub Actions enabling attackers to execute malicious code and extract sensitive information. Prominent attacks such as "S1ngularity" and "Shai-Hulud" have demonstrated the catastrophic potential of exploiting unsanitized pull request data to inject harmful code into CI pipelines, compromising entire ecosystems. JFrog's AI-powered security bot, RepoHunter, has been developed to proactively identify and report such vulnerabilities before they are exploited, successfully alerting maintainers across multiple open-source projects. The AI-assisted technology has been both a tool for protection and a weapon for attackers, as seen in a recent campaign where malicious bots targeted repositories from major organizations like Microsoft and DataDog. RepoHunter's efforts have led to the responsible disclosure and remediation of numerous vulnerabilities, mitigating risks to global financial systems, AI infrastructure, and countless applications. Despite these defenses, the persistence of CI vulnerabilities underscores the need for organizations to continuously audit and secure their CI/CD environments to prevent future supply chain attacks.

JFrog has achieved the status of Microsoft Solutions Partner with certified software for Azure, indicating that its Software Supply Chain Platform is technically validated by Microsoft for efficient operation on Azure. This designation reflects JFrog's commitment to cloud integration and operational performance standards set by Microsoft, as well as its proven ability to deliver effective solutions for organizations using Azure. JFrog's deep partnership with GitHub enhances this capability by offering seamless integration with GitHub-native workflows, which provides developers with real-time security feedback and vulnerability remediation. The platform's integrations with Azure services, such as Azure DevOps, Azure Private Link, and Azure Blob Storage, support a secure and resilient DevSecOps lifecycle. Enterprises across various industries have successfully migrated to JFrog on Azure to enhance their software supply chains, achieving high uptime, reduced infrastructure maintenance, and compliance with industry standards. This recognition underscores JFrog's dedication to providing a secure, scalable, and efficient foundation for modern software development and delivery on the Microsoft Cloud.

AI is revolutionizing the software landscape by generating an overwhelming number of digital artifacts, necessitating robust governance and security measures to manage this influx. JFrog is positioning itself as a pivotal player in this evolving ecosystem by providing a comprehensive platform that serves as the control plane for software supply chain integrity and governance at scale. With the rise of AI agents and the increasing generation of binaries, JFrog emphasizes the importance of a single system of record to ensure traceability, security, and compliance across all binaries, whether they are human-made or AI-generated. The company advocates for a proactive approach to governance, termed "DevGovOps," to automate and enforce security policies in the software development process. As AI continues to advance, JFrog's platform aims to mitigate risks associated with "shadow AI" by providing centralized governance and proactive security, ensuring that enterprises can maintain trust and control in a rapidly changing technological environment.