AI-Powered Attacks: A New Era of Software Supply Chain Risk for India

HYDERABAD, India – July 29, 2026 – The engine of modern software development, fueled by open-source components and collaborative coding, is facing a sophisticated and rapidly evolving threat. Artificial intelligence is no longer just a tool for innovation; it's being weaponized to transform software supply chain attacks from indiscriminate, large-scale campaigns into surgical, high-impact precision strikes. This new reality was the focus of a forum here today, where software supply chain leader Sonatype and research firm Forrester unveiled a stark warning for global enterprises, with a particular focus on India's burgeoning technology sector.

The joint research, detailed in the report "The Trust Economy of Software: How AI is Reshaping Software Supply Chain Risk for Financial Services," reveals a fundamental shift in attacker methodology. As enterprises embrace AI to accelerate development, malicious actors are exploiting the same technologies to target the very foundation of how software is built. The implications are profound for India, a global hub for financial technology, digital engineering, and Global Capability Centers (GCCs).

"AI is helping development teams assemble software faster, but it is also accelerating the number of decisions they make about what software to trust," said Abhishek Chauhan, Senior Director of Technology and India Country Head at Sonatype. "Attackers understand that shift. They are no longer relying only on scale; they are investing in precision attacks that look familiar, targeting developers directly, and executing before traditional controls have a chance to intervene."

The New Anatomy of an Attack: Precision Over Scale

The days of simple, opportunistic malware are fading. The new breed of software supply chain attack is targeted, deceptive, and designed to infiltrate projects at their inception. Sonatype's analysis of nearly 10,000 malicious package advisories from January 2020 to May 2026 paints a clear picture of this evolution.

The most alarming statistic is the explosion in targeted attacks. The number of malicious package advisories classified as precision campaigns skyrocketed a staggering 75-fold in just two years, jumping from 28 in 2023 to 1,576 in 2025. This isn't just more noise; it's a more intelligent and dangerous adversary.

These attacks rely on deception. The research found that 47% of malicious packages now actively impersonate trusted software, a technique known as brandjacking. Attackers use familiar names, logos, and descriptions to trick unsuspecting developers into incorporating malicious code into their applications. Furthermore, 53% of malicious packages analyzed in 2025 were designed to execute during the installation phase on a developer's machine, often using package installation hooks to run malicious scripts. This "shift-left" attack strategy ensures the compromise occurs before most traditional security scanners have a chance to inspect the code, effectively bypassing downstream defenses.

The sophistication of these packages is also increasing. More than one in four now use advanced techniques like code obfuscation, multi-stage droppers that download further malware, and hidden backdoors, signaling the involvement of more organized and well-funded threat actors.

India's Tech Hub in the Crosshairs

For India, these findings are not just an academic warning; they are a direct alert to a strategic economic risk. As one of the world's fastest-growing software development hubs, the nation's enterprises and vast network of GCCs are prime targets. The sheer volume of open-source software consumed and produced in India creates a massive attack surface.

The financial services sector is particularly vulnerable. According to Sonatype's earlier data from Q1 2025, financial services companies accounted for a staggering 66% of the open-source malware attacks the company blocked, highlighting the persistent focus of attackers on this high-value industry. As Indian fintech continues to innovate, its reliance on open-source components and rapid development cycles makes it a fertile ground for these new precision attacks.

Global Capability Centers, which handle critical software development and IT operations for multinational corporations, represent another critical point of risk. A successful supply chain attack on a GCC in Bangalore or Hyderabad could have immediate and devastating ripple effects across a parent company's global operations, compromising data and systems worldwide. The interconnected nature of these centers makes them a highly attractive target for adversaries seeking to breach major international corporations.

From Code to Crisis: Real-World Attack Vectors

The threats described in the report are not theoretical. They are manifested in real-world attack campaigns that are becoming more common. The impersonation tactic, for example, is seen in the constant stream of "typosquatting" packages discovered in public repositories like npm (for JavaScript) and PyPI (for Python). A developer might accidentally type react-domm instead of react-dom and unknowingly install a malicious library. With AI, attackers can now automate the creation and distribution of thousands of such deceptive packages, testing which ones are most likely to be downloaded.

The targeting of developer environments is equally concerning. By compromising a developer's machine, attackers can steal API keys, private credentials, and intellectual property, or use the machine as a launchpad to inject malicious code into a company's official source code repository. Advanced persistent threat (APT) groups, such as the notorious Lazarus Group, have been observed orchestrating industrialized malware campaigns specifically targeting developers, with Sonatype identifying over 800 Lazarus-associated packages in 2025 alone.

Building a Resilient Supply Chain: Beyond Reactive Security

In the face of these evolving threats, the old model of reactive vulnerability management—scanning for problems in production and patching them—is no longer sufficient. The consensus among security experts is the urgent need to shift to a proactive governance model.

This involves embracing industry frameworks like the National Institute of Standards and Technology (NIST) Secure Software Development Framework (SSDF) and the OWASP Software Supply Chain Security guidance. A core tenet of these frameworks is creating and maintaining a Software Bill of Materials (SBOM), which acts as a detailed inventory of all components in a piece of software, including open-source libraries and their transitive dependencies. An accurate SBOM provides the essential visibility needed to manage risk.

Enterprises must implement strong internal policies for component selection, using centralized repository managers to enforce rules about what software is approved for use. Automated policy enforcement is key, allowing security teams to block known malicious or vulnerable components from ever entering the development lifecycle. This approach moves security from a final-stage gatekeeper to an integrated, continuous part of the development process.

The Double-Edged Sword: AI as Both Weapon and Shield

The narrative is not entirely one-sided. While AI empowers attackers, it is also a critical tool for defenders. Security firms are increasingly leveraging AI for enhanced threat detection, using machine learning models to analyze vast datasets and spot anomalous patterns that could indicate a sophisticated attack. AI can automate vulnerability analysis, prioritize remediation efforts, and even suggest code fixes, helping security teams keep pace with development.

As organizations accelerate AI adoption, governance must evolve alongside development practices. This was a key point made by Ashutosh Sharma, VP and Principal Analyst at Forrester, during the event. He noted that software development is entering a new phase where understanding how software trust is established is paramount for balancing innovation with resilience.

For India's financial institutions and GCCs, the challenge is clear: lead the next phase of AI-driven development by governing software assembly with the same speed and discipline with which it is being created. In this new trust economy of software, confidence is not just about compliance; it is a strategic asset that enables faster delivery, stronger resilience, and the responsible adoption of transformative technologies.