IA · 29 July 2026 · 5 min read
The Race to Secure Agentic AI: Cyera Acquires Oasis Security in $1B Deal
In brief: Data security enterprise Cyera has signed a letter of intent to acquire cybersecurity startup Oasis Security for approximately $1 billion in a predominantly cash transaction. The acquisition aims to unify non-human identity management and autonomous AI agent security into a single platform, addressing the rapid expansion of automated bots across corporate software ecosystems.
by Team Mocchi's
A Billion-Dollar Deal for Non-Human Identities
The rapid rise of autonomous AI agents operating within corporate networks is fundamentally reshaping enterprise cybersecurity requirements. Highlighting this shift, data security firm Cyera has signed a letter of intent to acquire cybersecurity startup Oasis Security for approximately $1 billion, as reported by TechCrunch. The deal, expected to be settled primarily in cash with the remainder in Cyera stock, aims to build a unified platform for managing data protection alongside non-human digital identities.
Cyera, recently valued at $12 billion following a $600 million funding round and surpassing $150 million in annual recurring revenue (ARR), continues an aggressive acquisition strategy in the security domain. Founded in 2022, Oasis Security had previously raised about $195 million from investors including Accel, Craft Ventures, and Cyberstarts. The startup specializes in managing Non-Human Identities (NHIs)—ranging from API keys and service credentials to autonomous AI agents interacting with enterprise software.
Why AI Agents Redefine the Enterprise Perimeter
Until recently, Identity and Access Management (IAM) systems primarily focused on monitoring human employees and their end-user devices. The advent of agentic AI models has upended this paradigm. AI agents do not merely generate text or answer static queries; they execute multi-step business workflows by querying databases, triggering external APIs, altering system configurations, and taking autonomous operational decisions.
As detailed in an analysis by MIT Technology Review, agentic AI represents a full systems, governance, and orchestration challenge that extends far beyond LLM inference alone. When hundreds or thousands of autonomous agents operate in parallel, the credentials and privileges granted to them become prime targets for cyberattacks or unintended executions. Without centralized governance tracking which agent was authorized by whom and to what exact resources, the risk of data exfiltration or unintended system execution scales dramatically.
Moving Toward Unified Data and Bot Security Platforms
Cyera's acquisition highlights a broader consolidation trend within the enterprise software market: merging data security posture management (DSPM) with machine identity security. Prior to Oasis Security, Cyera acquired specialized startups such as Ryft and Genie Security to reinforce its posture against AI-weaponized threats and unchecked automated scripts.
Integrating Oasis Security’s capabilities will allow enterprise platform teams to monitor AI agent behavior in real time, applying policy-aware access controls that grant or revoke privileges based on real-time risk assessments. The explicit goal is to prevent autonomous agents from converting static API tokens and service accounts into unmonitored backdoors inside corporate networks.
Mocchi's take
For companies across Europe and Italy that are deploying AI agents into operational workflows—from automated customer support to complex internal management—this acquisition offers a clear lesson: agent performance cannot be decoupled from security architecture. Provisioning AI agents with broad API tokens or unchecked access to internal databases without dedicated non-human identity management exposes infrastructure to severe security risks. The opportunity for development and IT leaders lies in embedding zero-trust access, granular policy controls, and strict sandboxing into custom agentic workflows from day one, ensuring every AI agent operates with least-privilege permissions and full auditability.