AI News AI资讯 5h ago Updated 3h ago 更新于 3小时前 50

As AI agents become employees, NewCore emerges with $66M to give them identities 随着AI代理成为员工,NewCore获6600万美元融资以赋予其身份

NewCore exited stealth with a $66M seed round, valuing it at $300M. Focus: Identity and governance for AI agents in the enterprise. Founders from Dome9, Check Point, Unit 8200, and major telcos. Bets legacy identity platforms will break under AI agent scale/complexity. Positioning as a ground-up solution vs. Okta/Microsoft add-ons. 网络安全初创公司NewCore以3亿美元估值完成6600万美元种子轮融资。 旨在为企业大规模部署的AI代理提供统一的身份验证、治理与控制平台。 创始团队认为,为人类设计的15-20年旧身份系统无法应对AI代理带来的新复杂性。 核心理念是将AI代理视为与员工同等级的“一等公民”身份进行管理。 解决方案直接挑战Okta、Microsoft Entra等传统身份管理巨头。

75
Hot 热度
70
Quality 质量
70
Impact 影响力

Analysis 深度分析

TL;DR

  • NewCore exited stealth with a $66M seed round, valuing it at $300M.
  • Focus: Identity and governance for AI agents in the enterprise.
  • Founders from Dome9, Check Point, Unit 8200, and major telcos.
  • Bets legacy identity platforms will break under AI agent scale/complexity.
  • Positioning as a ground-up solution vs. Okta/Microsoft add-ons.

Key Data

Entity Key Info Data/Metrics
NewCore Seed Funding $66 million
NewCore Post-money Valuation $300 million
NewCore Lead Investor Cyberstarts
McKinsey AI Agent Deployment 25,000 agents alongside 60,000 employees
Goldman Sachs AI Agent Test Tested Devin as a "new employee"

Deep Analysis

NewCore’s pitch is a clean, almost inevitable problem statement: when your workforce includes digital entities that operate at machine speed and scale, the identity systems designed for human logins and 9-to-5 permissions are going to shatter. It’s not a speculative bet; it’s observing a reality in motion at firms like McKinsey. The real insight isn't that AI agents need identity—it's that their identity must be a fundamentally different class of entity from the start, not an afterthought bolted onto a legacy IAM (Identity and Access Management) stack.

The $300M pre-product valuation screams that investors are buying the timing and the team more than a proven platform. Zohar Alon’s experience selling Dome9 to Check Point is crucial here. He didn’t just found a company; he successfully navigated the brutal consolidation cycle of cybersecurity. He’s now playing a similar game: identifying a core infrastructure weakness (identity) right as a disruptive force (AI agents) is about to amplify that weakness exponentially. The "split-key" architecture mention, though cut off, hints at a technical moat—likely a way to cryptographically separate agent and human authority, solving a critical delegation-of-trust problem that current systems fumble.

The challenge is the elephant in the room: Okta and Microsoft. Their counter-narrative—"we’re just extending our platform"—is powerful because it’s cheaper and easier for CISOs to accept. NewCore’s counter-argument, that you can’t properly govern a jet engine by treating it as a fancier propeller, is technically sound but a harder sell. They’re fighting not just against competitors, but against institutional inertia and the sunk cost fallacy. The market window is narrow: they must become the default standard before the Okta/Microsoft "agentic patches" become "good enough" for the average enterprise.

This move also signals a broader shift. We are entering the age of "non-human entities" (NHEs) as first-class citizens in corporate networks. This isn't just about identity; it’s about auditability, liability, and real-time control. Who is accountable when an AI agent, acting on its permissions, causes a data breach? NewCore is building the answer key and the lock at the same time. Their success hinges on whether enterprises believe a new, purpose-built platform is a strategic necessity versus a costly convenience. The early, bleeding-edge adopters—those running thousands of agents—will be their proving ground. If they can demonstrate a tangible reduction in agent-related risk and administrative overhead there, the $300M valuation might look like a bargain.

Industry Insights

  1. The "Identity-as-a-Service" market will bifurcate into human-centric (legacy) and agent-centric (next-gen) stacks, creating new acquisition targets.
  2. AI agent "passports" will emerge, requiring embedded, auditable policies and real-time revocation mechanisms far beyond API keys.
  3. Enterprise security architectures will undergo foundational redesign to accommodate autonomous digital entities as permanent, privileged users.

FAQ

Q: Why can't companies just use Okta or Microsoft's existing identity tools for AI agents?
A: They can, but those are extensions of platforms designed for human workflows and lifecycles. NewCore argues this creates governance gaps and complexity at scale, like using a spreadsheet to manage an air traffic control system.

Q: What is the core innovation NewCore is proposing?
A: Treating AI agents as "first-class identities" from day one, not as service accounts. This implies a new data model and lifecycle management (creation, permissioning, monitoring, termination) built for machine-scale operations and autonomy.

Q: Does this matter to businesses not yet using AI agents heavily?
A: It’s a leading indicator. As AI agent adoption grows, identity governance will become a critical bottleneck and risk vector. Companies that plan their identity infrastructure now will avoid costly, chaotic migrations later.

TL;DR

  • 网络安全初创公司NewCore以3亿美元估值完成6600万美元种子轮融资。
  • 旨在为企业大规模部署的AI代理提供统一的身份验证、治理与控制平台。
  • 创始团队认为,为人类设计的15-20年旧身份系统无法应对AI代理带来的新复杂性。
  • 核心理念是将AI代理视为与员工同等级的“一等公民”身份进行管理。
  • 解决方案直接挑战Okta、Microsoft Entra等传统身份管理巨头。

核心数据

实体 关键信息 数据/指标
NewCore 种子轮融资金额 $66 million
NewCore 投后估值 $300 million
NewCore 主要投资方 Cyberstarts (领投), Index Ventures, Evolution Equity Partners
McKinsey 部署的AI代理数量 25,000个
McKinsey 员工数量 60,000名
Zohar Alon 创办的上一家公司 Dome9 (后被Check Point收购)

深度解读

NewCore的出现,精准地切中了企业AI应用从“工具化”迈向“代理化”的一个关键痛点:身份管理。我们正在见证一场静悄悄的革命,AI不再仅仅是写代码、做PPT的工具,而是像高盛测试的Devin那样,被当作“新员工”对待,甚至与人类员工协同办公(如麦肯锡)。当这些数字员工数量激增,它们的权限、行为审计和生命周期管理,就成了企业安全的新边疆。传统基于人类员工设计的身份系统(如Okta、Entra ID)在此显得捉襟见肘,因为它们的核心逻辑是为人服务的,无法原生承载海量、动态、且能力日益强大的AI代理身份。

这家公司最尖锐的洞察在于:它不是在旧体系上“打补丁”,而是直接定义了一个新战场。创始人Zohar Alon那句“15-20年的老平台会被打破”,不是危言耸听。传统巨头正试图通过添加“代理功能”来应对,但这本质上是给马车装上发动机,结构上依然是为人设计的。NewCore则宣称自己是“原生”为混合劳动力(人类+机器+AI)构建的,这就像云原生应用对比传统本地部署软件,架构差异决定了未来扩展性和安全性的上限。3亿美元的种子轮估值,表明资本市场相信身份管理这个本已庞大但停滞的市场,正被AI浪潮撕开一道新的、巨大的裂口。

时机选择也堪称精妙。此刻恰是企业从“试验AI代理”转向“规模化部署”的转折点。头痛还未全面爆发,但预见性已足够。创始团队组合(连续成功创业者、军方情报背景的CTO、大型企业CIO)提供了技术深度、安全洞察和商业化落地的多重保障。他们提出的“分离密钥”等架构,暗示着在AI代理的身份认证和控制上,需要全新的技术范式,而非简单套用OAuth或SAML。

更深层看,这揭示了AI时代企业软件的一个核心演进方向:从“增强人的能力”到“管理新的数字实体”。身份将成为AI代理的“操作系统”。谁掌握了这个新范式下的身份基础设施,谁就掌握了下一代企业软件栈的关键入口。NewCore赌的,正是Okta们未能及时转身的窗口期。

行业启示

  1. 身份基础设施将发生范式迁移:企业身份管理的核心对象将从“自然人”扩展至“自然人+AI代理”的混合实体,催生全新的技术架构和市场。
  2. 安全左移:AI治理前置化:对AI代理的权限控制、行为审计和合规性管理,必须在部署之初就作为核心身份能力内置,而非事后补救。
  3. “云原生”思维入侵安全领域:为新工作负载(AI代理)构建“原生”的身份平台,比在旧系统上兼容扩展具有颠覆性优势,这为初创公司创造了挑战巨头的机会。

FAQ

Q: NewCore要解决的核心问题究竟是什么?
A: 它致力于提供一个统一平台,用于大规模地认证、管理和控制与人类员工协同工作的AI代理,将其视为具有独立权限和生命周期的“一等公民”数字身份。

Q: 它与Okta、Microsoft Entra等现有身份管理巨头有何本质区别?
A: 关键区别在于架构设计出发点。传统平台基于人类员工设计,现为AI代理添加功能;而NewCore宣称从零开始,原生构建用于管理人类、机器与AI代理混合劳动力的身份系统。

Q: 为什么投资人此时给出如此高的种子轮估值?
A: 投资者押注于AI代理大规模企业应用是确定性趋势,而现有的身份管理基础设施存在明显短板。新市场窗口已现,且创始团队具备成功的连续创业与深厚行业经验。

Disclaimer: The above content is generated by AI and is for reference only. 免责声明:以上内容由 AI 生成,仅供参考。

Agent Agent 安全 安全 融资 融资
Share: 分享到: