독서리서치레이더투자 프레임워크
로그인 / 회원가입
← 읽기로 돌아가기
주제DisconfirmAI 편집부업데이트됨 2026-11-11 04:02

遗留系统阻碍AI落地

4 단계1 건5 출처

【背景】企业级AI落地正遭遇“创新税”:据Google Cloud风险与合规副总裁Jeanette Manfra于2026年11月发布的调研,49%的美国受访企业因本地部署的遗留软件推迟或取消了AI项目。这一现象并非孤例,早在2025年,Zapier与Deloitte的报告已指出集成困难是主要障碍。

核心矛盾在于零和博弈:IT团队被迫将工程周期用于维护老旧系统(如修补漏洞),而非转向AI现代化。Manfra认为这是供应商锁定导致的架构僵化,建议通过“安全设计”和AI驱动防御来解绑;而AWS等云服务商则强调,即便上云,若底层架构仍基于预AI假设(如静态容量、数据孤岛),依然无法支撑智能体工作负载。目前业界共识是遗留系统构成硬性瓶颈,但关于“彻底重构”还是“渐进式中间件集成”的路径选择尚无定论。

현재까지의 과정

  1. 2025-02-20

    CIO.com报道引用Ensono首席战略官Brian Klingbeil观点,指出遗留应用因数据存储限制和高昂维护成本,挤占了AI项目的预算与数据资源。

    cio.com ↗
  2. 2025-09-15

    Deloitte发布《2025 AI趋势》报告,显示近60%的AI领导者将“与遗留系统集成”列为采用代理型AI的首要挑战。

    deloitte.com ↗
  3. 2025-10-28

    Zapier发布调研,78%的企业在将AI工具与现有系统连接时遇到困难,53%认为集成过程极其困难。

    globenewswire.com ↗
  4. 2026-11-10

    Google Cloud副总裁Jeanette Manfra公布对1250名美国企业决策者的调研,确认49%的企业因遗留软件直接导致AI项目延迟、缩减或取消。

    cloud.google.com ↗
로그인 / 회원가입
독서리서치레이더투자 프레임워크

아직 해결되지 않은 문제

  • 企业在面对遗留系统时,应优先选择彻底重构代码,还是通过API/中间件进行渐进式集成?
  • 随着AI加速发现旧代码漏洞,传统IT运维团队是否有能力应对激增的安全补丁需求?
  • 供应商锁定(Vendor Lock-in)在多大程度上限制了企业选择最佳AI架构的自由度?

출처

  • cloud.google.comGoogle Cloud Blog: New survey finds enterprises trapped between legacy software and AI modernization↗
  • globenewswire.comGlobeNewswire: Zapier survey finds 4 in 5 enterprises struggling to integrate AI with legacy systems↗
  • deloitte.comDeloitte US: AI trends 2025: Adoption barriers and updated predictions↗
  • aws.amazon.comAWS Migration & Modernization: Why Legacy Infrastructure Blocks AI Success↗
  • cio.comCIO.com: Outdated business apps can cloud your AI vision↗

이 주제 커버리지

  • Jeanette Manfra, Vice President of Risk and Compliance at Google Cloud: Nearly Half of Responding Enterprises Postpone or Cancel AI Projects Due to Legacy Systems2026-11-10뉴스

    Based on a survey of 1,250 US business decision-makers, Jeanette Manfra, Vice President of Risk and Compliance at Google Cloud, argues that legacy software systems are acting as an 'innovation tax': 49% of respondents stated that on-premises software has actually delayed, reduced the scope of, or canceled AI projects, while 39% face licensing and negotiation constraints due to vendor lock-in. She proposes a zero-sum game mechanism—where IT engineering cycles are consumed by legacy maintenance, leaving no capacity for AI modernization—and recommends architectural choice freedom, secure-by-design foundations, and AI-driven defense.