The $686M Week That Defined AI Agent Security

What the platform does: NeuralTrust provides a unified platform for monitoring, governing, and securing enterprise AI agents. Its product suite includes:

• TrustGate — a secure AI gateway for model and tool-call control
• TrustGuard — runtime threat prevention against attacks, malicious activity, and hallucinations
• TrustLens — discovery and visibility into shadow AI agent sprawl
• TrustTest — automated red-teaming with a library of 150+ attack scenarios

Why it matters: NeuralTrust reports that it inspects millions of AI agent interactions daily, with about 1.2% requiring intervention . Its Q1 2026 annual recurring revenue (ARR) doubled its full-year 2025 ARR, signaling strong enterprise demand [user question, supported by funding context]. The company serves customers in banking (Abanca, Banc Sabadell), aviation (Iberia, Air Europa), and telecoms, and is ISO 27001 certified with recognition in Gartner's Agent 365 Report . The funding will accelerate engineering growth, product integration, and European market expansion .

Founded in 2022 by CEO Joan Vendrell, CTO Victor Garcia, and COO Alejandro Domingo, the company is headquartered in Barcelona with an office in London .

NewCore: Identity Infrastructure for the Agentic Workforce

Tel Aviv and San Francisco-based NewCore emerged from stealth on June 15, 2026, with $66 million in seed funding at a $300 million post-money valuation . The round was led by Cyberstarts, with participation from Index Ventures and Evolution Equity Partners, along with prominent angel investors including Assaf Rappaport and Yevgeny Dibrov .

What the platform does: NewCore describes itself as a security-first identity platform rebuilt from the ground up for the workforce that now includes humans, machines, and AI agents . Key differentiators include:

• AI agents treated as first-class identities with their own lifecycle, trust score, and revocation path — not as legacy service accounts
• A split-key architecture combined with visual multi-factor authentication and hardware-bound credentials
• Centralized discovery, governance, and control of every identity in the enterprise

Why it matters: Traditional identity and access management (IAM) systems were built for human employees. As enterprises deploy thousands of AI agents into production systems, those agents need their own authentication, permissions, and audit trails. NewCore's platform sits on top of existing IT infrastructure without requiring a forklift upgrade . The company has more than 10 design partners and plans to begin monetization in summer 2026 . Co-founders include Zohar Alon (former CEO of Dome9, acquired by Check Point for $175M) and Amihai Neiderman (Unit 8200 veteran) .

Cyera: The $12B Bet on Data Security for AI

Data security firm Cyera raised $600 million in a Series G round at a $12 billion valuation, led by Evolution Equity Partners with participation from Cyberstarts and Temasek . The round quadruples Cyera's valuation from 18 months earlier (when it was valued at $9 billion), and brings the company's total funding to over $2 billion . Existing investors including Accel, AT&T Ventures, Blackstone, Coatue, and Spark Capital participated . Customers include Paramount and Chipotle .

What the platform does: Cyera helps organizations track, classify, and secure their data assets across cloud and on-premises environments. The company frames its mission as building "the trust layer for the AI era" — ensuring that enterprises can safely use AI tools and agents without exposing sensitive data . Its AI Guardian platform provides visibility into which data AI systems can access and enforces policies to prevent unauthorized exposure.

Why it matters: Cyera's valuation acceleration — from $9B to $12B in roughly five months — reflects investor conviction that data security is the critical bottleneck for enterprise AI adoption . As AI agents gain access to corporate databases, customer records, and internal documents, knowing what data exists and who (or what) can touch it becomes essential.

Broader Trend: Agent Security as a Standalone Category

Taken together, these three deals — NeuralTrust, NewCore, and Cyera — illustrate three layers of a new cybersecurity stack for the agentic era:

The common thread: enterprises are moving AI agents from experimentation into production, and existing security tools — built for human users, static workloads, and known threat models — don't cover the gaps. Investors are placing large bets that the companies that build the security layer for autonomous AI will be the next CrowdStrikes and Okta.

BlackFog also joined the wave this week, launching ADX Vision for macOS on June 18, 2026, extending its shadow AI detection and data exfiltration prevention platform to Apple endpoints . ADX Vision operates as a native system extension, monitoring processes, clipboard events, and browser traffic in real-time to block sensitive data from reaching unsanctioned large language models . The move highlights that enterprise security teams are waking up to the risk of employees using unapproved AI tools on any device.

Bottom line: June 2026 will be remembered as the week AI agent security became a mainstream investment category. With nearly $700 million in fresh capital deployed across seed, growth, and late-stage rounds, the message is clear: the cybersecurity industry is building a new layer of defense for a world where AI agents are as common as human employees.

This article is based on confirmed source reporting. The Palo Alto Networks—Deutsche Telekom sovereign AI security partnership mentioned in the initial query could not be verified from the provided source set and is not included here.