Threat Labs Report: India 2026

Jump to a section

Key findings AI use

AI: Adoption and use trends AI: App use and data policy violations Most blocked AI apps

Agentic AI adoption

User adoption of AI agents Rising use of AI APIs outside the browser

Malware downloads Phishing Personal cloud apps risk

Personal apps activity Data policy violations in personal applications

Recommendations Netskope Threat Labs About this report

9 min read

Key findings

In this report, we examine how organizations’ digital estates are evolving under the influence of AI, and the security and data protection risks arising from this evolution. AI is now interlaced with the fabric of daily workflows, from agentic systems and SaaS APIs to the personal apps employees lean on when they think no one is watching.

The researchers have also examined how attackers are exploiting trusted cloud platforms to deliver malware inside Indian organizations, and why data policy violations are evolving in a manner that ought to prompt every software developer to take notice.

The shift from shadow AI: For a long time, shadow AI, where employees used personal AI accounts for work, was running rampant. However, we have seen a major shift occurring within Indian organizations over the past year, with the use of managed AI tools jumping from 30% to 77%. While the use of personal accounts dropped nearly by half in the same period, challenges remain. Nearly one in five users is still using both personal and enterprise accounts at work, indicating that governance alone is insufficient. Organizations aiming to eliminate shadow AI completely must remove the need for employees to use personal accounts by ensuring enterprise tools are just as frictionless as those people use at home.

Claude is beating ChatGPT: ChatGPT is still the king of the hill in India, and used in 88% of the organizations we track. But Anthropic’s Claude is making an incredible run, now sitting at 84% adoption, compared to a little over 30% a year ago. The real story, though, is in the plumbing. When we look at where developers are actually connecting their internal systems via APIs, Anthropic has already pulled ahead. 85% of organizations are plugged into Anthropic’s API, eclipsing OpenAI at 64%.

AI is ubiquitous: In India, 82% of workers are directly engaging with AI apps, 97% are using SaaS tools that use AI in the background, and 92% are using apps that leverage user data for training models. While AI adoption continues to rise, it is showing signs that it may be reaching a plateau.

Code is the new crown jewel at risk: Over the last 12 months, source code was involved in nearly half of all data policy violations related to AI use. As developers rush to use AI to debug and build, proprietary logic is being exposed at an alarming rate.

Cloud platforms remain a malware delivery channel: Attackers continue to abuse trusted cloud services to distribute malicious content, aiming to evade detection, and increase the likelihood of user interaction. Microsoft OneDrive and GitHub were most frequently targeted, with 12% and 9.5% of organizations detecting malicious content on those platforms respectively.

Personal applications continue to pose data-exposure challenges: Personal cloud and AI applications remain widely used in workplace environments. LinkedIn, ChatGPT, and Google Drive are the most commonly used personal applications, and regulated data, source code, and intellectual property are the types of sensitive data most often at risk of leaking through them. These trends highlight the importance of data loss prevention (DLP) controls, user awareness, and strong governance practices.

AI use

AI: Adoption and use trends

AI use across organizations in India has continued to increase over the past year, indicating increasing maturity and confidence as AI becomes more deeply embedded in business processes and day-to-day operations.

User adoption has also expanded significantly. The share of users actively using AI applications has increased from 55% to 82%, indicating not only greater availability of AI tools but also more consistent use across the workforce.

In parallel, organizations in India have made considerable progress in reducing shadow AI risk, by moving users away from personal AI accounts and toward organization-managed tools. Over the past year, the use of personal AI applications dropped from 79% to 41%, while adoption of organization-managed AI solutions increased sharply from 30% to 77%. At the same time, users switching between personal and enterprise accounts rose from 10% to 18%, indicating that organizations still need to balance strong governance with usability, accessibility, and feature parity, alongside tighter instance-level controls.

Overall, this transition reflects improved governance and oversight, as well as a broader shift toward managed environments that strengthen data protection and regulatory oversight, while still enabling innovation. But there is still room for improvement.

In India, the top AI applications follow a slightly different hierarchy than global trends. ChatGPT remains the most widely adopted AI application, used in 88% of organizations. Interestingly, Anthropic Claude has strengthened its position in second place at 84%, ahead of Google Gemini at 77%. This differs from the typical global ranking, where ChatGPT leads, followed by Google Gemini and then Anthropic Claude.

The following chart shows how the use of leading AI applications has evolved over the past year in India, highlighting marked changes in platform preference. ChatGPT has remained consistently strong throughout this period, maintaining its position as the most widely used application.

Anthropic Claude saw a sharp rise in adoption from December 2025, quickly narrowing the gap with the market leaders, and now reaching second place. Meanwhile, Google Gemini and Microsoft Copilot have remained relatively stable, with steady use patterns.

AI: App use and data policy violations

As AI adoption grows across organizations in India, concerns around data exposure are becoming increasingly important. AI tools are now commonly used for summarizing documents, generating reports, and supporting business processes that often entail the use of sensitive business or customer data, consequently expanding the potential attack surface.

In this context, data protection remains a top priority, particularly amid ongoing shadow AI risks.

Analysis of data policy violations in India shows that source code is now the category of data local employees most often put at risk, accounting for 49% of incidents. Regulated data and intellectual property each account for 23% of incidents, and passwords and API keys 5%. On average, organizations in India experience more than 3,000 AI-related data policy violations per month, underscoring the greater exposure of proprietary development assets within AI-driven workflows, and supporting the need for robust DLP controls and well-governed AI use. In response, organizations are strengthening their ability to monitor and control data flowing into AI applications, with a sustained increase in DLP deployments for AI, which increased from 42% to 52% over the past year.

Most blocked AI apps

Organizations in India are often blocking specific applications due to security, privacy, or compliance concerns. Although policies differ across organizations, certain tools are restricted more frequently than others, reflecting where perceived risk is highest.

ZeroGPT is the most frequently blocked AI application at 43%, followed by Sider AI at 40% and Indian application Kenyt.ai at 39%. Sider AI is commonly associated with browser-based AI assistant extensions integrated into web workflows, while Kenyt.ai is often used for conversational automation and customer engagement use cases. These tools can increase data exposure risk due to their deep integration into user activity and limited visibility into data treatment.

Overall, these patterns imply that organizations in India are not only responding to individual application risks but also strengthening broader governance frameworks to ensure AI use aligns with security and compliance requirements. These trends and policies reflect data observed throughout the past year.

Agentic AI adoption

User adoption of AI agents

AI adoption across organizations in India has continued to grow over the years, with capabilities increasingly embedded into both operational and customer-facing workflows.

Adoption now spans multiple layers: 82% of employees actively use AI applications directly, while 97% use applications that include AI-powered features. In addition, 92% interact with AI systems that leverage user data to train algorithms and models.

This shows how deeply AI is embedded in everyday workflows, often extending beyond explicit use, and into built-in functionality within widely used tools, and the extent to which sensitive data may be exposed through direct or indirect interaction with AI systems.

Rising use of AI APIs outside the browser

Even when AI agents and applications are deployed on-premises within Indian organizations, they generally rely on cloud-hosted models delivered via SaaS or enterprise AI platforms. These systems regularly communicate through API-based integrations rather than browser interfaces.

For example, browser-based interactions with OpenAI typically occur via chatgpt.com, while internal tools and AI agents connect programmatically through api.openai.com.

In India, Anthropic’s API (api.anthropic.com) has surpassed OpenAI in adoption, with 85% of organizations connecting to it, compared to 64% for api.openai.com and 36% for api.assemblyai.com. This shows a growing preference for alternative model providers in API-driven AI architectures.

Overall, dependence on API-based integrations accentuates the expanding role of embedded AI services in operational, administrative, and enterprise systems, where secure, well-governed connectivity is critical.

Malware downloads

Attackers frequently exploit cloud platforms to distribute malware, leveraging users’ trust in well-known and legitimate cloud services. While providers actively remove malicious content, even short delays in detection can allow successful infections and internal propagation.

Across organizations in India, Microsoft OneDrive and GitHub are among the most commonly abused platforms for malware distribution, affecting 12% and 9.5% of organizations respectively. This shows a broader shift in attacker behavior toward using legitimate cloud infrastructure rather than suspicious or unknown domains, and making malicious activity harder to detect within normal traffic patterns.

Phishing

Phishing remains one of the most successful techniques used by attackers, despite years of security awareness training and improvements in defensive controls. Attackers have steadily refined their methods, moving beyond simple credential-harvesting pages to techniques such as malicious OAuth applications, fake authentication portals, and reverse proxies that capture credentials and session tokens in real time.

Phishing susceptibility among users in India has improved significantly over the past year. Since August 2025, the number of users clicking on phishing links has dropped by almost 50%, from 115 to 55 per 10,000 users. This trend suggests that organizations are becoming more effective at reducing phishing risk through a combination of stronger email security controls, increased user awareness, and broader adoption of phishing-resistant security measures.

Even so, phishing remains a critical risk to organizations in India, especially organizations relying heavily on SaaS applications. Identity has become a primary target, with credentials, access tokens, and session cookies providing direct access to business-critical resources, rendering them extremely valuable to attackers seeking to gain a foothold in corporate environments. In India, Microsoft is by far the most commonly impersonated cloud brand, accounting for 53% of observed cloud phishing activity. Many of these campaigns are designed to compromise Microsoft 365 accounts or steal credentials that can be reused across multiple services. Apple is the second-most impersonated brand at 11%, with attackers frequently leveraging the company’s trusted reputation to increase the likelihood of user interaction.

These trends illustrate the continued significance of monitoring identity-related threats and strengthening protections around cloud accounts, authentication flows, and user access to critical business applications.

Personal cloud apps risk

Personal apps activity

Across India, the widespread use of personal cloud and online applications in workplace environments continues to blur the line between corporate and personal data use.

LinkedIn is the most widely used personal application at 90%, followed by personal ChatGPT accounts at 82%, and personal Google Drive accounts at 80%. While much of this activity is legitimate, supporting collaboration, productivity, and networking, it also introduces data security risks when sensitive information is involved.

From personal AI accounts to file-sharing and communication tools, these applications remain key points of possible data exposure, especially when used outside approved workflows or during employee transitions.

Data policy violations in personal applications

Organizations across India deploy a range of controls to reduce the risk of accidental data exposure through personal cloud and AI applications. These include DLP controls to monitor and control sensitive data flowing into these applications, blocking uploads to unmanaged apps, and offering real-time guidance to users regarding risky actions they are undertaking.

When employees attempt to upload sensitive data into personal applications, regulated data is the most exposed category, accounting for 33% of all data policy violations related to personal cloud apps, followed closely by source code and intellectual property at 30% each, and passwords and API keys at 6%.

Personal Google Drive accounts are triggering the most upload blocks (37%), followed by ChatGPT (33%) and Google Gmail at (30%).

This distribution shows a relatively balanced mix of compliance-sensitive and proprietary data in Indian environments, where multiple data types add to overall risk. Strengthening DLP coverage, boosting user awareness, and enforcing clear data-handling policies remain essential to reducing both insider and external risks.

Recommendations

With the growing use of AI tools, both managed and personal, and the misuse of personal cloud apps, it is essential to strengthen visibility, improve policies, and prioritize proactive defenses to protect your organization in this fast-changing threat landscape.

Based on the trends uncovered in this report, Netskope Threat Labs strongly encourages organizations across India to take a fresh look at their overall security stance:

Inspect all HTTP and HTTPS downloads, including all web and cloud traffic, to prevent malware from infiltrating your network. Netskope customers can configure their Netskope One NG-SWG with a threat protection policy that applies to downloads across all categories and all file types.
Block access to apps that do not serve any legitimate business purpose or pose a disproportionate risk to the organization. A good starting point is a policy to allow reputable apps currently in use while blocking all others.
Use DLP policies to detect potentially sensitive information, including source code, regulated data, passwords and keys, intellectual property, and encrypted data, being sent to personal app instances, AI apps, or other unauthorized locations.
Use Remote Browser Isolation (RBI) technology to provide additional protection when visiting websites in categories that may pose a higher risk, such as newly observed or newly registered domains.
Use Netskope One AI Gateway to gain visibility and control over AI applications and API interactions, helping secure data flows between users, applications, and LLMs.
Deploy Netskope One AI Guardrails to enforce consistent protections against sensitive data exposure, unsafe prompts, and policy violations across managed and unmanaged genAI environments.
Use Netskope One GenAI App Security to discover sanctioned and unsanctioned AI applications, apply real-time controls, and enforce governance policies across personal and enterprise AI use.
Leverage Netskope One AI Analytics to monitor AI adoption trends, user activities, and DLP incidents, facilitating organizations to better understand and reduce AI-related risk exposure.
Consider Netskope One AI Red Teaming to actively identify vulnerabilities and misconfigurations in private AI deployments before they may be exploited in production environments.

Netskope Threat Labs

Staffed by the industry’s foremost cloud threat and malware researchers, Netskope Threat Labs discovers, analyzes, and designs defenses against the latest cloud threats affecting enterprises. Our researchers are regular presenters and volunteers at top security conferences, including DEF CON, Black Hat, and RSA.

About this report

Netskope provides threat protection to millions of users globally. The information presented in this report is based on aggregate use data collected by the Netskope One platform for a subset of Netskope customers in India.

The statistics in this report are based on the period from April 1, 2025, through April 30, 2026. Stats reflect attacker tactics, user behavior, and organization policy.

Threat Labs Report: India 2026

Jump to a section

Key findings

AI use

AI: Adoption and use trends

AI: App use and data policy violations

Most blocked AI apps

Agentic AI adoption

User adoption of AI agents

​Rising use of AI APIs outside the browser

Malware downloads

Phishing

Personal cloud apps risk

Personal apps activity

Data policy violations in personal applications

Recommendations

Netskope Threat Labs

About this report

Rising use of AI APIs outside the browser