What does Cyera do for enterprise AI agents?

Cyera builds a unified platform that governs access for every human, machine, and AI agent across the enterprise. Its Agent Guardian and Cyera Endpoint products track what AI agents actually do between prompts and responses, including the tool calls, database queries, and actions in between. The Oasis Security acquisition added non-human identity management, so enterprises can connect where sensitive data lives with which agent identities can actually reach it.

Why did Goldman Sachs invest $400M in Cyera?

Goldman Sachs Alternatives structured this as a Series G extension because it sees AI agent security as one of the defining enterprise technology categories of the next decade. As agents multiply faster than security teams can track them, the gap between what they are trusted to do and what they can actually access becomes a critical governance problem. Goldman's managing director Irit Kahan cited Cyera's technology, customer traction, and execution as the factors that positioned it to become the platform enterprises rely on.

What is non-human identity management and why does it matter?

Non-human identity (NHI) management covers the credentials and service accounts used by machines and software agents rather than people. Traditional identity and access management was designed around human employees logging in. AI agents, however, operate autonomously using service identities, API keys, and OAuth tokens. If those identities are compromised or over-permissioned, an agent can access and exfiltrate data across systems without any human involvement. NHI management closes that gap by applying the same least-privilege discipline to machine identities that IAM applies to human ones.

How does the Cyera platform integrate with AI coding agents like Claude Code or Cursor?

Cyera Endpoint extends governance to the devices where local coding agents run their work. Agents like Claude Code and Cursor increasingly handle sensitive files, credentials, and configuration data on developer machines. Cyera tracks the tool calls and data accesses those agents make at the endpoint layer, giving security teams visibility into agent activity even when it takes place outside cloud environments.

What does Cyera's funding trajectory tell enterprise buyers?

Cyera has raised $1.4 billion in 2026 alone, across three rounds: $400M in January, $600M in June at a $12B valuation, and this $400M Goldman Sachs extension. That pace of capital formation reflects high enterprise demand, not just investor speculation. Every major existing investor participated in the Series B with a larger check. For enterprise buyers evaluating AI security platforms, that trajectory is a signal about long-term vendor stability and roadmap investment.

Cyera Raises $400M from Goldman Sachs for AI Agent Trust Layer

Cyera raised a $400 million extension to its Series G from Growth Equity at Goldman Sachs Alternatives on September 22, 2026, at an unchanged valuation of more than $12 billion. The round brings Cyera’s total 2026 fundraising to $1.4 billion and its all-time total to more than $2 billion. The company, founded in 2021 by CEO Yotam Segev and CTO Tamar Bar-Ilan, is building what it calls the trust layer for the agentic enterprise: a platform that unifies data security and non-human identity governance to control what AI agents can see and do.

What Cyera Is Building and Why It Matters

The enterprise AI security problem Cyera is addressing is structural. Most security architecture was designed around human employees and conventional applications: people log in with credentials, access systems, and generate audit trails. AI agents operate differently. They reason autonomously, assume service identities, call tools and APIs, query databases, and take actions at machine speed, often without a human approving each individual step.

That creates a gap. An AI agent can hold legitimate enterprise credentials, pass standard access controls, and still take actions that are misaligned with business intent, excessively broad in scope, or outright harmful, without any existing security control catching the problem until after it happens.

Cyera’s answer is a platform that brings two historically separate disciplines together: data security (knowing where sensitive information lives and how it flows) and identity governance (knowing which principals can reach it). By connecting those two layers, the company gives enterprises a unified way to answer the question their security teams keep asking: which agents can access what, and what are they actually doing with it?

The Product Architecture: Data, Identity, and Endpoints

Cyera’s platform has three core components, each addressing a different layer of the agent security problem.

Agent Guardian and Cyera Endpoint are the behavior-tracking layer. Unlike endpoint detection tools that look for known malware signatures, these products record what AI agents do between receiving a prompt and producing a response. That includes the specific tool calls made, the database queries executed, and the files or credentials accessed. The visibility extends from cloud environments down to developer workstations, where local coding agents like Claude Code and Cursor increasingly handle sensitive files and configuration data.

Oasis Security (acquired by Cyera in July 2026 for approximately $1 billion) provides non-human identity management. Oasis built a platform to discover, classify, and govern the service accounts, API keys, OAuth tokens, and machine identities that software and agents use. By integrating Oasis, Cyera can now connect a data classification (this is sensitive financial data) with an identity classification (this service account can reach it) in a single system.

The combination is what gives Cyera’s trust layer framing its substance. Knowing where sensitive data is, and knowing which identities can reach it, are only useful security controls when you can enforce them in the same place.

The 2026 Funding Trajectory

Round	Date	Amount	Lead	Valuation
Series F extension	January 2026	$400M	Undisclosed	~$9B
Series G	June 2026	$600M	Evolution Equity	$12B+
Series G extension	September 22, 2026	$400M	Goldman Sachs Alternatives	$12B+

The pace is notable. Three rounds totaling $1.4 billion inside nine months is unusual even in a well-funded category. The September round structure is also a signal: Goldman Sachs Alternatives, which runs a $17B+ growth equity book, chose to extend the June round at the same valuation rather than negotiate a new price. That framing says the firm wanted access at the known price more than it wanted to establish a new anchor, which is a different kind of conviction than leading a new round at a step-up.

Every major investor from the Series B also returned with a larger commitment, which Cyera’s leadership cited as evidence that investors with direct visibility into customer usage and renewal rates chose to increase their exposure.

Why Enterprise AI Leaders Should Pay Attention

The security gap Cyera is building against is not theoretical. Enterprises are deploying AI agents into production systems at a rate that outpaces the governance frameworks meant to control them. Agents that have standing access to HR systems, financial databases, and customer records present a fundamentally different risk profile than the SaaS applications those systems replaced.

Three dynamics are converging to make this urgent:

Scale: The number of non-human identities inside large enterprises already exceeds the number of human employees in many organizations. AI agents will accelerate that ratio dramatically. Without tooling purpose-built for machine identity lifecycle management, the effective attack surface for credential compromise grows faster than any team can manually track.

Action depth: Modern AI agents do not only retrieve data. They write, modify, trigger workflows, and issue API calls to downstream systems. A compromised or misaligned agent with write access is a substantially worse scenario than a compromised data read credential.

Regulatory exposure: Data protection regulations, particularly in financial services, healthcare, and government, hold enterprises accountable for unauthorized data access regardless of whether that access was taken by a human or an autonomous agent. The legal and compliance frameworks have not yet fully caught up to agentic AI, but the liability exposure exists now.

Cyera’s platform is designed to address all three. The Agent Guardian behavior tracking produces the audit trail regulators will eventually require. The data classification layer identifies the sensitive assets most at risk. The Oasis NHI layer governs the identities agents use to reach those assets.

For enterprise security and AI governance teams evaluating their stack, the investment signals a category that is maturing from a nice-to-have to a procurement requirement. The comparison point is useful: agent-specific security platforms such as Zenity and Hush Security have each raised significant rounds in 2026, and the AIUC certification standard launched in September reflects the same underlying pressure: enterprises need a way to verify that agents operate within defined boundaries.

Cyera’s differentiation is the data layer. Most agent security platforms focus on behavior monitoring and access controls at the identity level. Cyera comes from the data classification side and has extended into identity and behavior, which gives it a different starting point for conversations about compliance and data governance.

What Comes Next

The Goldman Sachs funding announcement outlined three uses for the capital: continued AI security product development, expansion into the U.S. federal market, and growth in EMEA and APAC.

The federal push is the most capital-intensive. Federal procurement requires FedRAMP authorization, cleared personnel, and compliance with a set of requirements that take significant time and money to satisfy before any agency contract can close. That Cyera is willing to invest in that pathway at this stage suggests it expects the federal demand for AI agent governance to be large enough to justify the cost.

The international expansion reflects a pattern visible across the AI security category: enterprise demand for agent oversight is not limited to U.S. markets, but deployment of adequate governance tooling is concentrated in early-adopter organizations. Reaching global enterprises before competitors establish those relationships is a window that narrows over the next 12 to 18 months as the category consolidates.

For enterprise teams building or scaling agentic AI deployments, the question Cyera is helping answer is not whether to govern AI agents but how to do it at scale without a separate governance stack for every agent workflow. If you are evaluating where AI fits into your operations, start with a conversation about your AI strategy here.

Sources: Cyera press release, September 22, 2026 | Goldman Sachs Alternatives announcement | Calcalist/Ctech reporting by Sophie Shulman | SecurityWeek coverage | The Next Web analysis