Keeper Security Report Finds Widespread Challenges in Managing Human and Machine Identities

Executive Summary

On May 6, 2026, Keeper Security released a new global research report titled "Identity Security at Machine Speed." The report, which surveyed 3,200 senior IT and cybersecurity leaders, paints a stark picture of the challenges modern enterprises face in managing digital identities. A key finding is that 89% of organizations are struggling with "identity sprawl"—a complex and rapidly growing landscape of both human and non-human identities (NHIs). The proliferation of Artificial Intelligence (AI) is identified as a primary accelerant of this problem. The report highlights significant operational and security gaps, including fragmented security tooling, delayed detection of credential abuse, and a lack of visibility into AI usage.

Key Findings

The report's findings are based on a survey of IT decision-makers across the U.S., Europe, Asia-Pacific, and the Middle East.

1. Identity Sprawl is a Universal Challenge

• 89% of senior IT leaders find it challenging to manage their organization's expanding identity footprint.
• This includes not just human users (employees, contractors) but also a growing number of non-human identities (NHIs), such as service accounts, API keys, and AI agents.

2. AI is Accelerating the Problem

• 43% of global respondents identified the management of AI-related NHIs as a top identity governance gap. This number rises to 51% in the United States.
• The use of AI is creating a massive number of new machine identities that require secure management, and most organizations are unprepared.
• 56% of leaders are concerned about employees inadvertently exposing sensitive corporate data to AI systems.

3. Visibility and Tooling Gaps

• 96% of respondents stated that disconnected or poorly integrated security tools create exploitable security gaps.
• 42% of organizations admit to a lack of visibility into "shadow AI"—the unsanctioned use of AI tools by employees.

4. Delayed Threat Detection

• A staggering 72% of organizations do not detect the misuse of credentials in real-time.
• Most incidents take hours to identify, with some taking days or even weeks, providing a large window of opportunity for attackers.

Impact Assessment

The trends identified in the Keeper Security report have significant implications for enterprise security:

• Increased Attack Surface: Every unmanaged human or non-human identity is a potential entry point for an attacker. Identity sprawl directly translates to an expanded and poorly understood attack surface.
• Credential-Based Attacks: The report's findings suggest that attackers who compromise credentials have a high probability of success, as detection is often delayed. This makes credential stuffing, phishing, and other identity-based attacks more potent.
• AI-Specific Risks: The lack of governance around AI and NHIs creates new risks. Compromised AI service accounts could lead to massive data breaches, and the feeding of sensitive data into public AI models can result in irreversible data leakage.
• Operational Inefficiency: Fragmented tooling not only creates security gaps but also leads to operational inefficiency, alert fatigue, and higher costs for security teams.

Compliance Guidance and Recommendations

The report implicitly calls for a shift towards a more unified and modern approach to identity and access management (IAM).

• Centralize Identity Management: Organizations should move away from siloed tools and adopt a centralized platform for managing all identities—human and non-human. This provides a single source of truth for visibility and control.
• Embrace Privileged Access Management (PAM): The principles of PAM—least privilege, zero trust, and robust auditing—should be applied to all identities, not just traditional admin accounts. This is especially critical for powerful NHIs.
• Improve Detection Capabilities: The 72% of organizations that cannot detect credential misuse in real-time need to invest in better monitoring and analytics. This includes user and entity behavior analytics (UEBA) and real-time alerting on suspicious identity activity.
• Develop AI Governance Policies: Organizations must establish clear policies for the acceptable use of AI. This includes creating an inventory of sanctioned AI tools, training employees on data handling, and implementing technical controls (e.g., via CASB or DLP) to prevent data leakage to AI platforms.