Logo
Become a Client

Salesforce在今年的开发者大会TDX上，正式发布了名为Headless 360的新举措，旨在使得Salesforce平台上的所有功能，均可通过API（应用程序接口）、MCP（模型上下文协议）服务器或CLI（命令行界面）命令的形式对外暴露，从而支持编程智能体或面向特定客户需求的自定义智能体进行调用。

Salesforce的这一动作，并不仅仅是一次产品能力开放，更反映出企业软件正在发生一个根本性变化：应用的核心对象，正在从“人”转向“AI”。当越来越多系统开始以开放接口、连接器以及Agent相关协议而非UI作为主要交互方式时，企业软件的竞争逻辑、产品形态与商业模式，也正在被重新定义。

从应用到Agent，竞争格局与商业模式重塑

企业软件市场正经历一场深刻变革。AI Agent不仅是技术升级，更是市场竞争、商业模式和生态体系的全面重置。IDC最新调研显示，2026年全球72%的企业已将AI Agent投入生产，51.6%已将Agent嵌入核心业务流程。Agent正成为企业软件的“新入口”，未来三年内，Agent接口将超过一半，传统接口将迅速降低接近至零。

传统软件厂商以功能和UI为核心竞争力，如今则被“结果导向、自动执行”的Agent所取代。Agent能够跨系统自动编排任务，推动业务流程从“人操作”转向“意图驱动、自动完成”。IDC预测，到2027年，Agent自动化将增强40%以上的企业应用能力，重塑三分之一的业务流程和工作流。

IDC的核心判断是：Agent并不是一次功能升级，而是在重构企业应用的形态。

应用正在“退居后台”：Agent成为新的执行层

在传统模式下，企业应用的价值建立在“界面+流程”之上。用户进入系统、触发操作、完成任务，软件围绕“人如何使用系统”来设计。

但Agent改变了这个前提。

它可以基于上下文理解需求、跨系统调用能力，并直接完成任务。在越来越多场景中，用户不再需要进入某一个具体系统，任务已经在后台被执行。

这意味着一个关键变化：应用不再是工作的入口，Agent才是。

一旦入口发生变化，应用的角色也随之改变——从“交互界面”转向“能力提供者”，从前台走向后台。

应用边界开始消失：竞争从产品走向生态

当Agent可以跨系统完成任务时，原本清晰的应用边界开始变得模糊。

同一个业务流程，可能同时调用CRM、ERP和供应链系统；而对用户来说，这一切被封装在一次“请求”之中。应用不再以单独系统的形式被感知，而是作为能力被调用。

关键影响在于：竞争逻辑正在发生改变——除了“谁的功能更强”，还需要考虑“谁更容易被调用和整合”。

这也是为什么当前市场中，开放接口、连接器以及Agent相关协议（如MCP、A2A）迅速升温。厂商不仅仅是构建产品，而需要争夺“进入Agent调用链”的位置。

如果无法进入这个链条，即使功能完整，也可能被绕过，逐渐边缘化。

定价体系被重写：从“使用权”到“结果价值”

相比产品形态的变化，更深远的影响正在商业模式层面显现。

过去几十年，企业软件的定价建立在“使用权”之上——按用户数、按模块、按许可收费。但在Agent模式下，这种逻辑开始失效。

一个Agent可以替代多个用户执行任务，自动化程度越高，企业获得的价值越大，但座席数量反而可能下降。

这带来一个根本性问题：当“使用量”不再等于“价值”，软件应该如何收费？

目前市场已经出现一些过渡模式，例如在座席基础上叠加Agent调用量或自动化流程计费，形成混合定价结构。

但更关键的变化在于：

软件行业正在从“为工具付费”，转向“为结果付费”。

而谁能够定义“结果”，并将其转化为可计量、可收费的单位，谁就有机会在这一轮变革中掌握价值分配权。Agent定价需简化，避免复杂结构成为创新和试点的障碍。

IDC观察：三条正在形成的主线

从当前市场演进来看，这一轮变化并不是单点突破，而是沿着几条清晰的主线展开。

首先，企业软件正在从“工具导向”走向“结果导向”。软件不再只是支持人完成工作，而是直接交付结果。

其次，集成逻辑正在从系统层上移到Agent层。过去复杂的系统集成，正在被Agent编排所替代。

更重要的是，竞争的核心正在从功能能力转向价值捕获能力。厂商之间的差距，不再只体现在“能做什么”，而在于“如何从结果中获得收入”。

谁会受益，谁面临风险？

在这一过程中，市场分化已经开始出现。

具备平台能力和生态控制力的厂商，更容易成为Agent的调度中心，从而掌握流量与价值入口。而那些依赖单点功能、定价僵化或生态封闭的厂商，则面临利润下滑和客户流失，被边缘化的风险——即使产品本身依然存在，Agent会绕过其产品，其不再处于用户路径之中。

换句话说，未来的竞争不只是“有没有Agent”，而是“是否在Agent体系中占据关键位置”。

结语：这不是一次技术升级，而是一次价值重构

Agent的快速普及，表面上看是AI能力在企业软件中的延伸，但其更深层的影响在于，它正在改变应用的形态、重塑竞争边界，并重新定义价值的获取方式。

这不是一次简单的技术升级，而是一场围绕“执行权”和“价值权”的重构。

对于软件厂商而言，真正的挑战不在于是否引入Agent，而在于——在一个由Agent主导的体系中，自己的产品究竟处于什么位置。Agent不仅是自动化工具，更是业务流程、行业应用和迁移的“新软件”，厂商可通过Agent快速扩展市场份额和收入。

与此同时，企业用户也需要重新评估自身的应用架构、自动化路径以及供应商选择策略，为Agent驱动流程、数据准备和跨应用编排做好组织准备，以确保在这一轮变革中持续获得业务价值与竞争优势。

IDC更多相关研究

如果您希望进一步了解Agent在企业软件中的落地路径、市场演进趋势或对自身业务的具体影响，欢迎与IDC分析师团队联系。IDC将基于持续的市场跟踪与研究，提供更具针对性的洞察与建议，支持企业与厂商在这一轮变革中做出更有前瞻性的决策。

请点击此处与我们联系。

Lizzie Li

Lizzie Li - Associate Research Director

Analyst Profile
Lizzie Li is Associate Research Director of IDC China's Enterprise System and Software Research that focuses on research and analysis of the China Datacenter, Cloud Computing, and IT infrastructure markets. She also provides intelligence and consulting services in customized projects for…

日本のテクノロジー市場で事業展開するベンダーやサービスプロバイダーにとって、AIインフラが成長するかどうかは既に決着した問題です。問題は、AI向けサーバーやストレージで構成される国内AIインフラ市場が1兆円の壁を超えようとする中で、どれだけ速く、どのような形で、そして誰がその市場を獲得できるかということです。

IDCの最新データと予測は、市場成長について明確な答えを示しています。以下は、このエコシステムのすべてのベンダーが理解すべき戦略的視点です。

1兆円への道筋：市場機会を定義する3つのマイルストーン

IDCは、今後5年間で国内AIインフラ市場が明確かつ力強い成長軌道を描くと予測しています。特に注目すべきマイルストーンは3つあります。

これらは楽観的な予測ではありません。政府の政策、エンタープライズのデジタル化への圧力、ハイパースケーラーのコミットメント、そしてAIの日本経済への不可逆的な統合という構造的な力に裏づけられており、その流れが反転する兆候はありません。

国内AIインフラの現在地：この成長を可能にする土台

これから起きることの規模を理解するには、日本のAIインフラ市場がいかに速く動いてきたかを把握する必要があります。2025年、国内市場は6,700億円に達しました。アクセラレーター搭載サーバーに限れば、2023年から2025年の3年間のCAGRは200%に迫り、世界平均を大きく上回っています。

この成長は偶然ではありません。政府の経済安全保障政策と国内資本の動員が意図的に重なり合った結果です。政府の技術的自立を推進するクラウド関連政策の下、国内資本のサービスプロバイダーや通信キャリアが、日本市場では稀に見るスピードで大規模なAIインフラ整備に乗り出しました。

物理的な変化も目覚ましいものがあります。かつての2U GPUサーバーから、ラックスケールシステムや水冷を前提とした複数ラック構成が標準的になり、データセンター全体がAIワークロードを中心に設計される時代へ向かっています。2025年の6,700億円規模の市場はこの土台の上に成り立っており、2030年の1兆円市場はその次に来るものの上に築かれます。

ベンダーが理解すべき市場の構造的ダイナミクス

2025年の国内AIインフラ市場は、ハイパースケーラーを含むサービスプロバイダーに大きく集中しており、市場支出全体の90.6%を占めています。この数字には、2030年に向けた競争環境を形成する3つのダイナミクスがあります。

・拡大を続けるハイパースケーラー： 投資シェアは2022年の39.8%から2025年には58.9%に急増し、3年間で19ポイントの上昇となりました。ハイパースケーラーのプラットフォームとロードマップへの対応は、この市場における存在感を維持するための前提条件であり続けます。

・戦略的重要性を持つ国内サービスプロバイダー： 政策支援を受けた国内サービスプロバイダーや通信キャリアを含むその他のサービスプロバイダーは、ハイパースケーラーの急拡大にもかかわらず、2025年のシェアを31.6%と維持し、2022年の31.8%からほぼ横ばいです。このセグメントは、純粋な価格競争よりもローカルな信頼、法規制対応、継続的なパートナーシップを重視する、強靭な顧客基盤です。

・次の成長フロンティアとなるエンタープライズの直接投資：現在9.4%という市場シェアながら、エンタープライズのAIインフラ直接投資は2022年比で絶対額が倍以上に拡大しており、成長の萌芽は確実に生まれています。現時点では大多数のエンタープライズが生成AIサービスやSaaSを通じてAIを利用していますが、AIへの野心が深まり、消費から保有へと移行するにつれて直接投資は加速します。今エンタープライズとの関係を構築するベンダーが、この波を最も有利な立場で捉えられます。

セミソブリンAIモデル：日本を特徴づける戦略的アーキテクチャ

2026年4月、マイクロソフトは2026年から2029年にかけて日本へ約1.6兆円を投資する計画を発表しました。これと合わせて、国内パートナー2社が国内で運用するAIインフラをAzureから利用可能にする構想が示されました。国内サービスプロバイダーの保有するAIインフラが、グローバルなハイパースケーラーのサービスレイヤーに接続されます。

IDCはこの構造を「セミソブリンAI」と捉えており、日本のAIインフラ戦略の特徴的なモデルとして急速に確立されつつあります。外国資本のハイパースケーラーへの完全依存でもなく、完全独立の国内AIインフラという過大なコストを強いるものでもない、現実的かつ政治的にも持続可能な折衷案です。

ベンダーにとって、このモデルは制約ではなく構造的な機会です。セミソブリンAIは、インフラ設計、システムインテグレーション、マネージドサービス、コンプライアンス対応、そして日本固有のAIプラットフォーム開発において、豊かで拡大し続ける市場を生み出します。このモデルを深く理解し、その中に意欲的に自社を位置づけるベンダーやインテグレーターが、2030年以降の日本AIインフラ市場の競争環境を定義することになります。

ベンダーへの示唆：動くなら今

国内AIインフラの基盤を築いた政策主導のアプローチは、需要主導の成長段階へと移行しています。もはや、適切なAIインフラが存在するかどうかは中心的な質問にはなりません。問うべきは「誰がエンタープライズのAI活用を、測定可能なビジネス価値を生む形へスケールさせるのか」です。この市場で競争するベンダーには、3つの行動指針が求められます。

• エンタープライズのエンゲージメントの加速：現在9.4%というエンタープライズのシェアが、明日の成長ストーリーになります。エンタープライズとの関係構築、日本特有のユースケース開発、ROI実証フレームワークに投資するベンダーが、この10年で最大の需要の波に乗る準備を整えられます。

• セミソブリンAIモデルへの適合： 国内のAIインフラ所有者、ハイパースケーラーのサービスレイヤー、政府の政策フレームワークの相互作用を理解することは、持っていたほうが良い背景知識ではありません。この市場で勝つための戦略地図です。

• 規模だけではない市場理解： 国内市場は持続的な地域へのコミットメント、深い技術的専門性に加え、日本の商習慣にも理解を示すベンダーを評価します。1兆円のチャンスは単なる取引量だけでは捉えられません。

ベンダーシェアや需要構造の詳細はIDC Worldwide Quarterly AI Infrastructure Trackerで継続的に分析しています。

関連する調査やご相談について

より詳細なインサイトや市場動向については、当社アナリストへお気軽にご相談ください。

Shinya Kato - Senior Research Manager, Enterprise Infrastructure, Data & Analytics, - IDC Japan

Analyst Profile
Shinya Kato is a Senior Research Manager at IDC Japan and is responsible for the data analysis and forecasting team of Japan enterprise infrastructure market. He analyzes the impact of product technology, service offerings, and marketing strategies on enterprise infrastructure market and provides market forecasts, focusing on the domestic enterprise storage systems market. Through understanding technology adoption trends, he also provides insight into emerging devices such as flash, accelerators, and quantum computing. In addition to researching the HPC and AI infrastructure markets, he is also investigating new consumption models such as Hardware-as-a-Service, to help stimulate the market. Prior to joining IDC, he spent more than 10 years at Silicon Graphics, which was later acquired by HPE, where he held various domestic positions in sales, marketing, and business development. He has covered a wide range of businesses, from infrastructure hardware and container-based data center facilities to digital asset management, industrial virtual reality, and software for media & entertainment. He also served as a product manager for enterprise internet security software and appliances at the emerging vendor. He holds a Bachelor of Economics degree from Rikkyo University.

2025年中国企业级MaaS市场经历了从试点到规模化应用的关键转折。无论是Token调用量还是实际营收，各个统计口径都呈现出倍数级的增长，同时Token消耗的快速攀升正在重新定义企业使用人工智能的方式。然而，在这一轮高速增长的背后，市场仍然面临性能、安全合规、回答质量等多重实际约束。IDC认为，MaaS厂商的竞争焦点正在从过去单纯的价格比拼，转向“价格、性能与工具链支持”的综合能力竞争。

市场总体规模：调用量增长16倍，但需理性看待基数效应

2024年，中国企业级MaaS市场按调用量统计的规模仅为114万亿Tokens，而到2025年，这一数字跃升至1944万亿Tokens，同比增长约16倍。在营收口径下，2025年中国公有云MaaS市场的规模达到30.7亿元人民币。与此同时，面向传统政企客户的大模型私有化部署市场也保持活跃，尤其在政务、金融、能源等领域形成了独立的采购与交付体系。

IDC预计2026年全年Token消耗量约为40,000万亿次，较2025年进一步增长约20倍。这一加速增长的主要驱动因素，是多模态大模型的逐步成熟以及Agent类应用的规模化落地。换句话说，市场正在从“文本生成”向“多模态理解与自动执行”扩展，每一次交互所消耗的Token量级也相应大幅提升。

Token消耗快速增长，但不同场景之间差异巨大

Token消耗的整体走势非常明确：2025年1月全市场日均消耗约为1.6万亿Tokens，到2025年12月底，这一数字已经攀升至9.6万亿Tokens。随着多模态大模型的进一步成熟，日均Token消耗的增长曲线将会变得更为陡峭。

不过，不同应用场景之间的Token消耗量级存在巨大差异，这一点往往被笼统的市场总量数字所掩盖。例如，一家投资机构在合同摘要场景中，一次处理几十个合同文档，单次消耗的Tokens可达200万。一家教育机构在其在线培训场景中，高峰使用期日均消耗高达2000亿Tokens。而使用多模态产品生成一个视频，也会消耗上千万Tokens。这些差异意味着，企业在评估MaaS服务时，不能只看单位Token的报价，还必须结合自身场景的实际消耗特征来判断总成本。

从应用场景的分布来看，当前公有云上MaaS的主要应用群体集中在泛互联网行业（游戏、娱乐、教育）、智能办公赛道、智能硬件赛道（智能汽车、手机、智能眼镜）以及大消费赛道。主要的应用场景包括角色扮演、短剧生成、市场营销、搜索、数据处理、数据分析和文档处理。而在传统政企的私有化部署项目中，应用场景相对收敛，主要集中在智能办公、数据处理与分析、市场营销等几个方向。

竞争格局：公有云头部集中，私有化部署市场更为分散

在公有云MaaS市场，按调用量计算，2025年火山引擎占据了接近一半的份额，其次是阿里云、百度智能云、硅基流动以及移动云。其他值得关注的厂商还包括腾讯云、商汤科技、华为云、天翼云等。

如果按营收口径来看，火山引擎仍然占据40%以上的市场份额，阿里云、百度智能云、智谱以及移动云位列前五。不同厂商在单位Token定价、折扣策略以及高价值场景的占比上存在显著差异。例如，某些厂商可能在低单价、高调用量的场景中占优，而另一些厂商则在高单价、专业场景中获得了更高的营收贡献。

在私有化部署市场上，格局则完全不同。传统政企客户出于数据安全、合规可控等考虑，仍然将私有化部署作为第一选择，也因此培育了众多的大模型平台私有化厂商。2025年IDC追踪到的头部厂商包括百度智能云、商汤科技、电信AI、中关村科金、创新奇智、星环科技以及中国电子云。相比公有云市场，私有化部署市场的集中度更低，这与政企采购的区域性、行业性特征有直接关系。

企业落地大模型时，性能、合规与质量优先于成本

高速增长的数字背后，企业在大模型落地过程中仍然面临一系列现实挑战。根据IDC的调研数据，影响大模型落地的Top5因素依次是：模型性能、安全合规要求、回答质量、在AI平台可用性以及成本效益。其中，性能指的是企业部署上线后的稳定性、并发数和可靠性；安全合规既要求模型在中国市场可以合规使用，也要求模型生成的内容本身合规；回答质量则直接决定了模型能否被大规模上线应用。

值得注意的是，成本效益在当前阶段排在第五位，并非企业最优先考虑的障碍。这一结果在一定程度上说明，企业当前更关注“模型是否能用、是否安全、是否稳定”，在这些前提条件满足之后，才会进入精细化的成本比较。然而，原文也明确提示了一种中长期风险：随着大模型应用场景的持续渗透，以及算力紧缺状况的延续，大模型的投资预算终将成为关键挑战。换句话说，成本问题目前还没有爆发，但它迟早会来。

MaaS市场的竞争要素正在发生结构性的变化

随着Token经济的兴起和市场参与者的迅速增加，MaaS市场的竞争规则也在发生变化。IDC认为，未来制胜的核心要素将集中在三个方面：价格与成本优化、性能、以及工具链与应用开发的支持。

在价格与成本方面，企业Token的日均消耗正在快速增长，尤其是各类Agent类产品的出现，进一步放大了Token的使用规模。随着规模化智能的到来，厂商需要关注的不是单纯降低单位Token的价格，而是帮助用户降低整体成本，同时确保输出的是高质量、有效用的Tokens。

在性能方面，大量行业场景对并发数和吞吐量有很高要求，与此同时算力紧缺的现状短期内难以根本缓解。能够通过底层算力优化来提升模型训练和推理效率的厂商，将在用户选型中获得明显的优势。

在工具链与应用开发支持方面，过去两年的MaaS市场中，买方选型的首要因素几乎都是价格，而且由于大模型迭代速度太快，买方通常选择直接调用大模型API，而不进行二次调优。但未来，企业将有更多垂直场景落地，这些场景离不开智能体的开发。因此，MaaS平台在工具链（如调试、评测、智能体编排）上的完备性，在用户选型中的重要性将不断提高。

展望：2026年预计20倍Token增长，但需注意情景依赖

基于现有数据和趋势，IDC对MaaS市场给出了明确的预测：2026年中国MaaS市场的Token消耗量将达到约40,000万亿次，按营收规模计算将达到约186亿元人民币。2024年至2030年的年复合增长率约为1154.9%，但这一数字是基于高增长情景得出的。

需要特别提示的是，如此高的复合增长率高度依赖于一系列前提条件：多模态模型的持续成熟、Agent类应用的大规模落地、算力供给不出现严重瓶颈，以及合规政策保持稳定。上述任何一个条件发生变化，实际增速都可能显著低于高增长情景。对于行业参与者和企业买方而言，在看到巨大市场机遇的同时，也有必要对风险保持清醒的认知。

对厂商与买方的建议

基于上述分析和IDC的调研数据，我们可以对MaaS厂商和企业买方分别提出几点建议。

对于MaaS厂商而言，在价格竞争的同时，应当优先解决高并发场景下的性能稳定性和合规推理能力。工具链能力（调试、评测、智能体编排）正在成为差异化竞争的关键，不可忽视。此外，可以考虑在垂直场景（如合同处理、视频生成、教育陪练）中建立深度优化能力，而不是在所有场景中采取同质化的竞争策略。

对于企业买方而言，在当前阶段应当优先验证模型的回答质量、安全合规性和并发稳定性，成本可以放在次优级的位置上。在选择MaaS厂商时，建议关注其工具链的成熟度，而不仅仅是API的单次调用价格。从长远来看，建议尽早建立Token成本的评估机制，避免业务规模扩张后出现成本失控的局面。

关于数据：本文所有数据和图表均来自IDC《中国AI软件市场半年度追踪，2025H2》。调用量口径统计企业通过公有云MaaS平台调用大模型API产生的Token数量；营收口径统计企业实际支付的公有云MaaS服务费用（不含私有化部署）；私有化部署市场统计面向政企客户的大模型平台软件及相关授权收入。

进一步交流

MaaS市场的爆发才刚刚开始，IDC将持续追踪这一市场的格局变化与技术演进。如需获取完整报告、订阅后续研究，或针对特定厂商与行业进行定制化分析，欢迎与IDC中国人工智能研究团队联系。

请点击此处与我们联系。

Yanxia Lu

Yanxia Lu - Research Director

Analyst Profile
  Yanxia Lu is a research director, focusing on big data and artificial intelligence (AI). Her responsibilities include big data information management platform, and big data analytics and applications. She is also involved in research on AI technology and enterprise…

The foundational question for any vendor or service provider operating in Japan’s technology market today is no longer whether AI infrastructure will grow. That is settled. The question is how fast, in what form, and most importantly who will capture the value as Japan’s AI infrastructure market, consisting of servers and storage for AI, approaches and surpasses the ¥1 trillion threshold.

IDC’s latest data and forecasts provide a clear-eyed answer. What follows is the strategic view every vendor in this ecosystem needs to understand.

The road to ¥1 trillion: Three milestones that define the opportunity

IDC projects Japan’s AI infrastructure market will follow a clear and compelling trajectory over the next five years. Three milestones stand out:

These are not optimistic projections. They are grounded in structural forces – government policy, enterprise digitalization pressure, hyperscaler commitments, and the irreversible integration of AI into Japan’s economic fabric – that show no signs of reversing.

Where we are today: The foundation that makes this possible

To appreciate the scale of what lies ahead, it helps to understand how quickly Japan’s AI infrastructure market has already moved. In 2025, the domestic market reached ¥670 billion. For accelerator-equipped servers alone, the three-year CAGR from 2023 to 2025 approached 200%, significantly outpacing the global average.

This growth was not accidental. It was driven by a deliberate convergence of government economic security policy and domestic capital mobilization. Under cloud-related policy frameworks designed to advance Japan’s technological sovereignty, domestically capitalized service providers and telecommunications carriers launched large-scale AI infrastructure buildouts at a pace rarely seen in the Japanese market.

Physically, the transformation is equally striking. Rack-scale systems and multi-rack liquid-cooled configurations are becoming the new standard. What was once a 2U GPU server is evolving into an infrastructure architecture where entire data centers are purpose-designed around AI workloads. The ¥670 billion market of 2025 is built on this foundation, and the ¥1 trillion market of 2030 will be built on what comes next.

The structural dynamics vendors must understand

Japan’s AI infrastructure market in 2025 is heavily concentrated in service providers, who accounted for 90.6% of total market spend. With that figure, three dynamics will shape the competitive landscape through 2030:

  • Hyperscalers will continue to expand. Their investment share surged from 39.8% in 2022 to 58.9% in 2025, a 19-point gain in three years. Alignment with hyperscaler platforms and roadmaps will remain a prerequisite for relevance in this market.
  • Domestic service providers will remain strategically critical. Other service providers including policy-backed Japanese service providers and telcos held 31.6% share in 2025, essentially unchanged from 31.8% in 2022 despite the hyperscaler surge. This is a resilient customer segment that values local trust, regulatory compliance, and sustained partnership over pure price competition.
  • Enterprise direct investment is the next growth frontier. At 9.4% of the market today, enterprise direct AI infrastructure investment has more than doubled in absolute terms since 2022, yet it remains nascent. Most large enterprises currently consume AI through generative AI services and SaaS platforms. As enterprises deepen their AI ambitions and move from consumption to ownership, direct infrastructure investment will accelerate. Vendors who build enterprise relationships now will be best positioned to capture this wave.

The semi-sovereign AI model: Japan’s defining strategic architecture

In April 2026, Microsoft announced plans to invest approximately ¥1.6 trillion in Japan between 2026 and 2029. Alongside this commitment came a plan to make AI infrastructure operated domestically by two partner companies available through Microsoft Azure, domestically built and owned infrastructure, connected to a global hyperscaler’s service layer.

This is what IDC describes as “Semi-Sovereign AI” and it is rapidly emerging as the defining model for AI infrastructure strategy in Japan. It represents a pragmatic and politically viable middle path: neither complete dependency on foreign hyperscalers, nor the prohibitive cost of fully independent domestic AI capability.

For vendors, this model is not a constraint. It is a structural opportunity. Semi-Sovereign AI creates a rich and expanding set of market roles in infrastructure design, systems integration, managed services, compliance, and the development of Japan-specific AI platforms. The vendors and integrators who understand this model deeply, and position themselves within it deliberately, will define the competitive landscape of Japan’s AI infrastructure market through 2030 and beyond.

What this means for vendors: Act now, not later

The policy-driven supply wave that built Japan’s AI infrastructure base is transitioning into a demand-driven growth phase. The central question is no longer whether the infrastructure exists, it does. The question is who helps enterprises put it to work, at scale, in ways that generate measurable business value. Three imperatives stand out for vendors competing in this market:

  • Accelerate enterprise engagement. The 9.4% enterprise share of today is the growth story of tomorrow. Vendors who invest in enterprise relationships, Japan-specific use cases, and ROI demonstration frameworks now will be positioned to ride the most significant demand wave of the decade.
  • Align with the Semi-Sovereign AI model. Understanding the interplay between domestic infrastructure owners, hyperscaler service layers, and government policy frameworks is not optional background knowledge. It is the strategic map for winning in this market.
  • Build for depth, not just scale. Japan’s market rewards vendors who demonstrate sustained local commitment, deep technical expertise, and an understanding of Japanese enterprise culture. The ¥1 trillion opportunity will not be captured by volume alone.

Ongoing vendor share and demand analysis is available via the IDC Worldwide Quarterly AI Infrastructure Tracker. For more detailed insights and market trends, please contact our analysts by completing this form IDC | Identifying Market Opportunities – Contact Us.

Shinya Kato - Senior Research Manager, Enterprise Infrastructure, Data & Analytics, - IDC Japan

Analyst Profile
Shinya Kato is a Senior Research Manager at IDC Japan and is responsible for the data analysis and forecasting team of Japan enterprise infrastructure market. He analyzes the impact of product technology, service offerings, and marketing strategies on enterprise infrastructure market and provides market forecasts, focusing on the domestic enterprise storage systems market. Through understanding technology adoption trends, he also provides insight into emerging devices such as flash, accelerators, and quantum computing. In addition to researching the HPC and AI infrastructure markets, he is also investigating new consumption models such as Hardware-as-a-Service, to help stimulate the market. Prior to joining IDC, he spent more than 10 years at Silicon Graphics, which was later acquired by HPE, where he held various domestic positions in sales, marketing, and business development. He has covered a wide range of businesses, from infrastructure hardware and container-based data center facilities to digital asset management, industrial virtual reality, and software for media & entertainment. He also served as a product manager for enterprise internet security software and appliances at the emerging vendor. He holds a Bachelor of Economics degree from Rikkyo University.

The Middle East War is no longer just a regional risk for technology leaders. It is becoming a global economic stress test — affecting energy prices, supply chains, inflation, business confidence, and IT budget decisions across Asia Pacific.

For CIOs, CFOs, and technology suppliers, the question is not simply whether IT spending will continue to grow in 2026. The sharper question is: which investments will still earn approval when volatility hits the budget?

That is the central signal from IDC’s latest webinar, Asia Pacific IT Spending Outlook 2026: Where to Win Amid Market Volatility. The market is not freezing. It is filtering.

This matters because the disruption is emerging against a market with strong underlying momentum. The opportunity is still there. The rules for winning it are changing.

The Middle East War is the shock

IDC’s earlier point of view on the Middle East War’s impact on IT spending outlined the broader global pressure points, from energy price volatility to supply chain disruption and cloud resiliency.

In Asia Pacific, those pressures are showing up in practical ways. Higher oil and electricity costs can raise operating expenses. Supply chain disruption can increase hardware costs and delay availability. Currency and pricing pressure can make approvals harder. As uncertainty rises, organizations move faster into contingency planning.

The first effect is not always a budget cut. More often, it is hesitation. Organizations start asking which projects are necessary, which can be phased, and which need to prove value faster.

As I noted during the webinar: “IT spending is likely to remain more resilient than in previous downturns because AI investment continues to support both the IT industry and broader economic growth.”

Resilience matters. It keeps the outlook from becoming a slowdown story. But resilience does not remove risk. The longer energy prices and supply chain disruption remain elevated, the more likely selective delays become targeted cuts.

Budget rotation is the signal

The most important shift is not how much organizations spend, but what they are willing to defend.

In a more volatile environment, buyers are becoming more risk-averse. ROI thresholds are rising. Procurement cycles involve more stakeholders. Vendors are seeing longer deal cycles, greater pricing pressure, and less predictable pipelines.

The visual above captures the core market signal: APAC IT spend is rotating, not disappearing.

Large transformations, experimental initiatives, and hardware-heavy refreshes are more likely to be deferred or reduced unless they are directly tied to productivity gains. Enterprise platform upgrades, GenAI pilots, and data and analytics programs are still in play, but they face tighter scrutiny and phased rollouts.

What remains protected? Investments tied to measurable outcomes. AI with clear ROI. Cloud optimization that supports flexibility and cost control. Cybersecurity that reduces risk and supports regulatory compliance. Infrastructure resiliency that helps organizations absorb disruption.

My colleague Vinayaka Venkatesh, summarized the shift directly: “The deals are not disappearing, but they are taking longer and require more effort to close.”

For technology suppliers, that line is the market cue. In 2026, vendors will not win by selling transformation in broad terms. They will win by helping buyers make the case for action now.

AI remains resilient, but proof matters more

AI remains central to the Asia Pacific technology agenda. IDC’s Asia/Pacific AI and GenAI spending outlook shows continued growth momentum as organizations invest in automation, productivity, and new business models.

But resilience does not mean every AI project gets a free pass.

One of the more important nuances from the webinar is the gap between AI infrastructure investment and enterprise AI execution. Hyperscalers and service providers continue to invest aggressively in AI infrastructure. At the same time, some enterprise AI initiatives are being de-scoped, delayed, or challenged on measurable outcomes.

This is where the “year of reckoning” for AI becomes real. AI use cases tied to cost savings, automation, customer experience, risk reduction, and operational efficiency will be easier to defend. Weak GenAI pilots, unclear operating models, or projects without a credible path to production will face more scrutiny from CFOs and cross-functional decision-makers.

AI for AI’s sake will not win the next budget cycle. AI with measurable business impact will.

What to remember for H2 2026 planning

As technology leaders plan for the second half of 2026, three takeaways matter most:

1. APAC IT spending is being disrupted, not derailed
The Middle East War is creating real pressure through energy costs, inflation, supply chain disruption, and weaker business visibility. But priority investments continue to support the market, especially AI infrastructure, cybersecurity, resiliency, and cloud optimization.

2. Budgets are rotating toward measurable value
Lower-priority initiatives, long-payback projects, and unclear business cases are facing delays. Investments that reduce risk, improve efficiency, support compliance, or deliver near-term ROI are better positioned for approval.

3. AI remains resilient, but the proof bar is rising
AI investment remains strong, but enterprise projects need clearer outcomes, stronger business cases, and a faster path from pilot to production.

The path forward is not about spending more for the sake of momentum. It is about knowing where to spend, what to protect, and what to pause.

Register for the on-demand webinar

The Middle East War is the shock. Budget rotation is the signal. Measurable value is the next move.

Register for the on-demand webinar, Asia Pacific IT Spending Outlook 2026 to see IDC’s latest scenario analysis, buyer data, and guidance on where IT spending is holding, where it is slowing, and what it means for H2 2026 planning.

Stephen Minton - Group Vice President, Data & Analytics - IDC

Analyst Profile
Stephen Minton is a group vice president with the IDC Data & Analytics group, focusing on ICT spending and macroeconomics. Mr. Minton is responsible for Worldwide ICT Spending programs, including the Worldwide Black Book, Worldwide 3rd Platform Spending Guides, and Worldwide Telecom Services Tracker. Mr. Minton's research expertise includes global ICT and economic analysis, and he tracks market data across hardware, software, services, telecom and emerging technologies. He is the author of papers that focus on the economic impact of IT, and is a regular speaker on the subject of IT spending. In 2002 he addressed the United Nations in New York, speaking to UN ambassadors on the subject of the Information Society. Mr. Minton previously worked with Digital Equipment Corporation (DEC), before joining IDC in 1998. Originally from Hartlepool in the North of England, he graduated from the University of Salford in 1995. He has also worked in the field of consumer market research with Millward Brown International.
Vinayaka Venkatesh

Vinayaka Venkatesh - Senior Market Analyst

Analyst Profile
Vinayaka Venkatesh is a senior market analyst for the IDC Asia/Pacific IT Spending Team based in Bangalore, India. He currently works as a vertical analyst, covering multiple spending guides for Asia/Pacific (excluding Japan and China) (APEJC). He joined IDC in…

レガシーシステムが稼働し続けるたびに、競合他社が優位を築いていく。日本の2.1兆円規模のITモダナイゼーション市場は待ってくれない—変革を急ぐ企業も同様である。

主要指標

1,304億円 — ITモダナイゼーションサービス市場規模（2025年）

10.2% — 年平均成長率（2025〜2030年）

2,123億円 — 市場規模予測（2030年）

約80% — 依然としてレガシーシステムを稼働させている大企業・中堅企業の割合

なぜ日本は世界を上回るペースで成長しているのか

日本のITサービス市場は2024年から2029年にかけて年平均6.6%成長すると予測されており、世界平均の3.6%のほぼ2倍にあたる。その背景には構造的な要因がある。日本は特有の重いレガシー資産を抱えている——長年にわたる汎用機（メインフレーム）やオフィスコンピュータなどへの投資、複雑な個別開発システム、そしてそれらを長年維持してきた人材がある。今、これら三つに起因する課題が重なり合う中、ITモダナイゼーションが避けられないものになっている。

富士通メインフレームのサポート終了

2022年、富士通はメインフレームおよびUNIXサーバー製品の販売・サポートの2030年前後の終了を発表した。この発表により、1,000社以上の企業が後戻りのできないカウントダウンに入り、日本市場全体でITモダナイゼーションの取り組みが加速している。

AIへの対応という至上命題

AIの活用には、緊密に統合されたデータパイプラインと近代的なビジネスプロセス基盤が前提となる——まさにレガシーシステムはこれらの実現を阻む要素となっている。AI競争力を維持したい企業にとって、ITモダナイゼーションはもはや選択肢ではない。

人口動態の圧力

日本のレガシーシステムを構築・維持してきた世代のエンジニアが退職しつつある。そのノウハウや技術が失われてしまう前に、知識とインフラを移行できる時間は着実に縮まっている。

モダナイゼーションへの三つのアプローチ

IDCはITモダナイゼーションサービスを三つの実行タイプに分類しており、それぞれがサービス企業に異なる意味をもたらす。

リホスト

既存のアプリケーション資産を維持しながら、レガシー以外のプラットフォームへリフト＆シフトする。予算や移行期間に制約を抱える企業にとっての入口となる手法である。

リライト

ビジネスロジックを変えずに、レガシーのソースコードを現代的な言語に変換する。管理された変革のための中間的なアプローチである。

リビルド

プロセス、データモデル、アーキテクチャをゼロから再定義する。最も高い価値をもたらす一方、最も複雑なアプローチでもある。

短期的には、リホストはリビルドに次ぐ2番目に大きなセグメントであり、メインフレームなどのEOL（End of Life）に対し早急な対応を要するに企業による支出が市場を牽引している——ただし既に成熟期を迎えており、今後はマイナス成長が予測されている。中長期的な成長機会は、アプリケーションのモダナイゼーション——リライト、リファクタリング、マイクロサービス化やクラウドネイティブアーキテクチャの採用——にある。

国内ITモダナイゼーションサービス市場 支出額予測：　2025年～2030年

Source: IDC Japan, 2/2026

企業がサービスプロバイダーに本当に求めているもの

IDCの調査では、レガシー依存度が相対的に高い大企業・中堅企業は、単なる技術的な実行だけを求めているのではなく、変革のパートナーを求めていることがわかった。セキュリティは基本的な前提として期待される一方、上位のニーズにはビジネスプロセス変革の支援やクラウド活用支援が挙がっている。

需要のシグナルはセクターによっても明確に異なる。

金融サービス

クラウドネイティブなアプリケーション開発能力、すなわち近代的なインフラ上で素早くイノベーションを起こす能力を優先している。

製造・流通

ビジネスプロセスの変革を優先している。基盤となる技術を刷新するだけでなく、業務に効率性とインテリジェンスを組み込むことを重視している。

全セクターを通じて、IDCは企業の期待に一貫した変化を観察している。ビジネス上の成果が主要な購買基準になりつつある。技術的な能力は当然のこととして見なされ、価値の創出が差別化要因となっている。

今、勝てるポジションを築くために

サービス企業にとって、競争上の必要性は明確だ。この市場で勝利する最良のポジションにある企業は、次の三つを実行する。

1. レガシーモダナイゼーションの実績を体系化する

過去の案件は活用されていない資産だ。サービス企業は、達成したビジネス成果——コスト削減、リードタイムの改善、AI対応力の解放——を体系的にまとめた資料を構築し、これを市場への訴求の核にすべきである。

2. AIの時代に向けた業種別のリファレンスアーキテクチャを開発する

汎用的なモダナイゼーションの提案は説得力を失いつつある。企業は自社のセクター、規制環境、そしてAIへの志向に合わせたシステムアーキテクチャと実装ロードマップを求めている。

3. 需要に先行してアプリケーションモダナイゼーション能力に投資する

リホストの波は既にピークに差し掛かりつつある。高い利益率をもたらす機会——リライト、リファクタリング、リビルド——がその後に続いている。クラウドネイティブとマイクロサービスの深い能力を培ったサービス企業こそが、2030年に向け企業から選ばれる存在となる。

IDCが提供するレポートのご紹介

IDCでは、国内ITモダナイゼーション市場の動向を詳細に分析したレポートを発行しています。

2026年 国内ITモダナイゼーション市場動向
 

本調査レポートは、IDCの国内サービス市場予測における主要な成長促進要因の一つであるレガシーシステム（老朽化・陳腐化、肥大化・複雑化、ブラックボックス化したシステム）のITモダナイゼーションについて、市場規模の中期予測を示すと共に、国内企業（ITバイヤー）の取り組み動向や、それを支援するサービスベンダーの動向を分析しています。国内ITモダナイゼーションサービス市場予測では、サービスセグメント別、実行タイプ別（リホスト、リライト、リビルド）、システムタイプ別、産業分野別に予測しています。これらの分析から、国内企業のITモダナイゼーション支援におけるニーズ変化や市場機会、サービスベンダーの支援サービスの特徴や戦略を包括的に把握できます。

関連する調査やご相談について

より詳細なインサイトや市場動向については、当社アナリストへお気軽にご相談ください。

Masaru Muramatsu - Senior Research Analyst, Software, Services, and IT Spending, IDC Japan - IDC Japan

Analyst Profile LinkedIn
Masaru Muramatsu is a senior research analyst, responsible for research and analysis of the Japanese IT services market, including IT consulting, systems integration, business services. Prior to joining IDC, Masaru worked to help digitalize local government in Japan, implementing software as a service (SaaS) in the education and taxation sectors. He also acquired experience in domestic and international sales/marketing with his work for a company that provided materials for electronic devices like smartphones, PCs, and printers. Masaru Muramatsu earned a master’s degree in engineering from Chuo University, Japan.

For years, digital accessibility, the practice of ensuring that digital products and services can be perceived, understood, and used by everyone regardless of ability, was treated as a compliance checkbox. That framing is no longer adequate. AI is reshaping accessibility into a strategic capability, one that is adaptive, continuous, and embedded in how people work, interact, and innovate.

As AI-enabled work becomes the norm, accessibility is no longer about supporting a small subset of users. It is about ensuring that everyone, across physical, sensory, cognitive, and neurodiverse dimensions, can fully participate in increasingly digital and AI-mediated environments. In this context, accessibility becomes foundational to productivity, inclusion, and ultimately business performance. Accessibility is also part of company culture: involving disabled and neurodiverse individuals in co-design, not just testing, creates more robust and adaptable systems. Sustaining long-term impact also requires investment in skills and culture, training employees, fostering inclusive design practices, and making accessibility a shared responsibility across teams.

The opportunity: AI as a scaler of inclusion and innovation

AI introduces a powerful opportunity to rethink accessibility at scale.

First, it enables real-time content adaptation. Capabilities such as automatic captioning, transcription, translation, and alternative text generation allow organizations to dynamically tailor content to different user needs. AI can also adjust reading levels, restructure complex information, and personalize interaction styles, supporting a broader range of cognitive and sensory preferences.

Second, AI supports continuous accessibility operations. Traditionally, accessibility has relied on periodic audits and remediation efforts. AI-driven testing tools now allow organizations to embed accessibility checks directly into development pipelines, transforming accessibility into a continuous, iterative process aligned with DevOps cycles.

Third, AI helps democratize innovation. By making tools and workflows more accessible, organizations can engage a wider and more diverse talent pool, including neurodiverse individuals and those historically underserved by traditional work environments. This expands creative input, improves problem-solving, and strengthens organizational resilience.

Finally, AI enables data-driven accessibility insights. Organizations can use AI to analyze accessibility barriers, monitor usage patterns, and measure outcomes, linking accessibility directly to business metrics such as productivity, employee engagement, and customer satisfaction.

Taken together, these capabilities shift accessibility from a cost center to a driver of innovation and competitive differentiation.

The pitfalls: Bias, complexity, and the risk of scaling barriers

Despite its promise, AI also introduces significant risks that organizations must actively manage.

One of the most critical challenges is bias in AI models. Many AI systems are trained on data and designed by teams that lack diversity. This can result in outputs that unintentionally exclude or disadvantage certain groups, particularly people with disabilities or non-standard interaction patterns. Without deliberate inclusion in design and testing, AI can reinforce existing barriers or create entirely new ones. Feedback loops that combine AI-driven insights with real user experiences are essential to countering this risk.

Another risk lies in inaccessible AI-generated content. While generative AI can produce fluent and polished outputs, these may still fail accessibility standards through improper structure, missing semantic cues, or formats that are difficult for assistive technologies to interpret. Auto-generated captions, for example, are often not accurate enough for compliance purposes.

The rise of agentic AI systems (autonomous AI that acts across workflows and applications without direct human instruction at each step) adds further complexity. If poorly designed, they can propagate inaccessible processes at scale, embedding friction into core operations rather than eliminating it.

There is also a governance challenge. As AI becomes embedded across systems, organizations must ensure clear accountability, transparency, and control over how accessibility preferences are handled, how decisions are made, and how user data is used.

Perhaps most importantly, over-reliance on automation can create a false sense of security. AI can scale testing and detection, but human validation, especially by people with disabilities, remains essential to identifying real-world usability issues.

Recommendations: Turning intent into impact

Organizations that want to lead in AI-enabled accessibility should focus on four key actions:

  • Prioritize accessibility as a design principle. Move from reactive compliance to proactive, accessible-by-design systems embedded in AI-enabled platforms and services.
  • Establish proactive AI accessibility governance. Integrate accessibility into AI governance frameworks early, ensuring inclusive workflows and avoiding costly retrofits.
  • Design for workforce adaptability and inclusion. Extend accessibility strategies beyond compliance to support diverse employee needs, including neurodiversity, aging workforces, and varying cognitive styles.
  • Act early to mitigate risk and maximize value. Early investment reduces remediation costs, strengthens trust, and positions accessibility as a strategic differentiator rather than a regulatory burden.

AI is redefining digital accessibility as a core element of how organizations operate, innovate, and compete. Those that embrace accessibility as a strategic priority will not only meet regulatory requirements but also unlock broader talent, improve user experiences, and build more resilient AI systems.

Erica Spinoni

Erica Spinoni - Senior Research Analyst, Worldwide AI-Enabled Future of Work & EMEA Practice Co-Lead

Analyst Profile
Erica Spinoni is a Senior Research Analyst for IDC’s Worldwide AI Enabled Future of Work practice, where she also contributes with regional expertise on EMEA-specific trends and dynamics. Her research helps technology vendors understand how emerging technologies reshape workforce practices…
Amy Loomis, Ph.D.

Amy Loomis, Ph.D. - Group Vice President, Workplace Solutions

Analyst Profile
Amy Loomis is Group Vice President for IDC’s worldwide Workplace Solutions.  Amy leads a team of analysts focused on the evolving nature of human resources, skills development, collaboration, and leadership across the employee lifecycle. Her research into the Future of…
Melinda-Carol Ballou

Melinda-Carol Ballou - Research Director, AI Assurance, ALM, Quality & Portfolio Strategies

Analyst Profile
Melinda Ballou delivers insights into the future of AI assurance, the impact of AI, ML and agentic adoption on agile and digital work, resilience, quality, product and software engineering, the role of technology in business and culture, and the evolution…

If you’ve spent the last few years talking to enterprise IT buyers about cost efficiency, you weren’t wrong. That was the conversation. But over the past few months, things have clearly shifted.

The outbreak of war in the Middle East, with its direct impact on people and organizations in the region, as well as broader effects on energy costs and IT manufacturing supply chains, is a primary driver. At the same time, early AI buildout pressures on memory supply were already raising concerns.

Today, when CIOs and their teams make technology decisions, the question is no longer, “How do we optimize spend?” It’s, “How do we keep the business running when things break?”

This shift shows up clearly in data from two major surveys on IT priorities and spending plans conducted in February and again in March. Concerns about hardware supply constraints have increased by more than 15%, and geopolitical risk is rising quickly. Meanwhile, traditional cost pressures, while still present, are starting to take a back seat.

This is not because cost no longer matters. It is because cost is now seen as downstream. If systems go down, supply chains stall, or cyber incidents escalate, cost becomes secondary very quickly.

What are IT buyers most concerned about in 2026?

When you talk to IT leaders today, the tone is different. There is more urgency, more realism, and more skepticism. They are thinking about exposure:

  • Where are we too dependent on a single cloud region?
  • What happens if a supplier cannot deliver?
  • How quickly can we recover from a cyber event?

Increasingly, they recognize that these risks are interconnected. A geopolitical event can disrupt supply chains, which impacts infrastructure, which affects applications, and ultimately hits revenue.

That is why IDC is seeing a clear pivot toward resilience.

Cybersecurity has moved to the top of the investment list globally, not just as a defensive measure but as a core part of keeping operations running. At the same time, organizations are accelerating investments in multi-region cloud architectures and backup strategies. Cloud security and multi-region resilience are now leading priorities across every major region.

IDC is also hearing from CIOs about a growing push to reduce dependency. CEOs are placing more focus on diversifying suppliers across all parts of the business. CIOs are responding by exploring sovereign cloud options and rethinking how and where infrastructure is deployed.

AI has not disappeared from the agenda, but it is being reframed. It is no longer just about innovation. It is about using automation and intelligence to keep systems stable under pressure.

Put simply, IT buyers are trying to build systems that can bend without breaking.

What does this shift mean for IT suppliers?

For suppliers, this shift creates both risk and opportunity.

The biggest risk is continuing to sell the way you did before. Leading with performance benchmarks, cost savings, or incremental features will not resonate the same way.

The opportunity is much bigger. Buyers are actively looking for partners who can help them navigate uncertainty. They are asking tougher questions:

  • What happens if this service goes down in one region?
  • How quickly can workloads move?
  • Where are the hidden dependencies?
  • How exposed am I if conditions worsen?

If you can answer these questions clearly and credibly, you move from being a vendor to becoming a strategic partner.

How should IT suppliers respond to rising resilience demands?

The challenge is that resilience means something different depending on where you sit in the ecosystem. The common thread is this: you must show how your offering performs under stress, not just under ideal conditions.

Cloud providers: How to prove resilience beyond scale

For cloud providers, this is a moment to rethink the narrative.

Scale and efficiency still matter, but they are no longer enough. CIOs want to know how your platform behaves when a region is disrupted, connectivity is constrained, or workloads need to move quickly.

This means making multi-region resilience the default, not an add-on. It also requires transparency about risk exposure and greater flexibility around sovereignty and localization.

In short, you are not just selling capacity anymore. You are selling survivability.

SaaS providers: Why continuity is now a core differentiator

SaaS providers are increasingly part of the critical path of operations. If your application goes down, the business feels it immediately.

Buyers want reassurance. They want to understand your disaster recovery posture, regional architecture, and dependencies. They want to know how their data is protected and how quickly services can be restored.

The vendors that stand out will clearly articulate how they maintain continuity, not just deliver functionality.

IT and professional services firms: From transformation to readiness

For services firms, the conversation has shifted from long-term transformation to immediate readiness.

Clients still care about transformation, but right now they need help answering urgent questions: Where are we exposed? What should we fix first? How do we prepare for multiple scenarios?

There is a real opportunity to lead with practical, actionable support. Rapid assessments, scenario planning, and resilience design are where clients need help now.

Speed matters. Clarity matters even more.

Communications providers: Why network resilience is now critical infrastructure

Connectivity has always been important. Now it is critical infrastructure in the truest sense.

Organizations are looking for redundancy, alternative routing, and, in some cases, entirely new connectivity models, including satellite and hybrid networks.

The differentiator is reliability under pressure. If you can demonstrate that your network keeps people and systems connected when other options fail, that becomes a powerful advantage.

Infrastructure vendors: Delivering certainty in uncertain supply chains

Hardware vendors are facing a different kind of scrutiny.

Availability and certainty in delivery are becoming as important as performance. Buyers want to know not just what the system can do, but whether they can actually get it, deploy it, and rely on it.

Transparency into supply chains, flexibility in configurations, and the ability to adapt to constraints are becoming key differentiators. In this environment, certainty is value.

Why IT buying decisions are shifting from optimization to assurance

Stepping back, what we are seeing is a shift in how technology decisions are made.

It is less about optimization and more about assurance. Less about peak performance and more about consistent operation.

The suppliers that win over the next six months will be the ones that can answer a simple but critical question:

What happens when things do not go according to plan?

From an enterprise IT leader’s perspective, that is no longer a hypothetical. It is the reality they are planning for every day. Resilience is no longer just a capability. It is the basis for trust.

What should IT suppliers do next?

If you are an IT supplier, now is the time to recalibrate how you engage with customers.

Start by pressure-testing your value proposition:

  • Can you clearly articulate how your offering performs under disruption?
  • Can you quantify how you improve resilience, not just efficiency?
  • Can you help customers understand and reduce their exposure?

Just as importantly, ground your strategy in real buyer insight.

IDC’s latest Future Enterprise Resiliency & Spending Survey (March 2026, Wave 2) provides a detailed view into how enterprise IT leaders across regions are reprioritizing risk, resilience, and investment decisions in response to geopolitical and supply chain disruption.

We encourage you to explore the survey findings to better understand:

Suppliers that align early with these shifts will be better positioned to engage, differentiate, and win. Because in this market, insight isn’t just helpful.

It’s your competitive edge.

Rick Villars

Rick Villars - Group Vice President Worldwide Research

Analyst Profile
Rick is IDC's leading analyst guiding research on the future of the IT Industry. He coordinates all IDC research related to the impact of Cloud and the shift to digital business models across infrastructure, platforms, software, and services. He helps…

AI is not just changing job descriptions; it is actively rewiring how work is coordinated, controlled, and created, and it is doing so on multiple fronts at once, inside the same organization.

AI Is Transforming Work on Multiple Fronts Simultaneously

Some of our IDC Future of Work predictions bring this into sharp focus: by 2027, 40% of current job roles in large organizations will be redefined or eliminated, accelerated by GenAI adoption. At the same time, by 2030, around 70% of new job roles in Europe are expected to be directly enabled by AI technology. This is not a neat “old jobs out, new jobs in” swap. It is a systemic reconfiguration of how value flows through the enterprise. Yet most leadership frameworks still present AI scenarios as if they were mutually exclusive: automate to cut headcount, augment to boost productivity, redesign work for agility, or push toward autonomous operations.

When Automation, Augmentation, and Autonomy Collide

On the ground, those dynamics do not arrive one by one; they collide. In the same business unit, you may be cutting FTEs as routine tasks are automated and taken over by “digital colleagues,” while simultaneously hiring AI orchestrators, prompt engineers, and automation product owners to keep up with demand for AI-adjacent skills. You may be tearing up long-standing workflows as agentic systems reshape a significant share of knowledge work, at the same time as parts of your operation drift toward near-autonomous execution, powered by employees building personal agents and conversational workflows that quietly absorb whole segments of the process. These are not options on a slide; they are concurrent forces acting on the same organizational fabric. Treating them like menu choices is not workforce planning. It is misdiagnosing an organizational phase transition, a fundamental shift in the underlying architecture of how work happens.

From Role-Based Models to Capability-Based Architectures

The uncomfortable truth is that many leaders are still planning for roles, new and “to be eliminated,” while AI is reshaping the landscape at the level of capabilities and architecture. You can see the tension in three simple signals. A clear majority of European organizations have already deployed or are piloting automation to offset chronic labor shortages. A growing share of executives openly discusses replacing positions with automation, and many plan to substitute a measurable portion of their workforce with “digital colleagues.” Meanwhile, by the end of this year, a meaningful slice of frustrated knowledge workers with no formal development background will be building their own agentic workflows to change how they work, regardless of what HR’s role catalog says. When people can spin up an agent in a week, any static role taxonomy you publish today is out of date tomorrow. The center of gravity moves from “what roles do we have?” to “what capabilities can we compose, and how fluidly can we recombine them as AI matures?”

Why Traditional Role Models No Longer Hold

Role-centric models allow for some seriously wrong assumptions: that tasks are stable enough to bundle into jobs, that jobs are stable enough to plan around for three to five years, and that hierarchies are stable enough to govern how value flows. Agentic AI quietly breaks all three. Tasks fragment, recombine, and migrate between humans and machines in near real time. Work starts to look less like a tidy org chart and more like a living graph of capabilities, human, machine, and hybrid. In that context, planning headcount against static job descriptions is like trying to architect a cloud-native platform using only server rack diagrams.

Architecture Determines the ROI of AI

However, IDC’s Future of Work research also shows that when enterprises invest in digital adoption and automated learning technologies, they can unlock substantial productivity gains. The pattern across these findings is consistent: it is the architecture that determines the yield of AI, not just the tools themselves. If your workflows are fragmented, AI struggles to “see” the end-to-end journey it needs to transform. When critical data is locked in legacy systems, it cannot provide the rich, contextual recommendations you were promised. When governance is tuned for stability rather than experimentation, it throttles the learning cycles AI needs to be useful. Layer on top the reality that many organizations openly acknowledge they lack the capability support to implement automation effectively, and a clear picture emerges.

AI Amplifies Existing Organizational Weaknesses

In that environment, throwing more AI at the problem does not fix anything. It amplifies what is already there. Bad processes simply run faster. Poor decisions scale further. Shadow automation blooms in the gaps, as frustrated employees script around the constraints of the operating model. AI becomes an accelerant, not a cure.

Reframing the Strategic Question for Leaders

This is why the strategic question has to change. Instead of asking, “Which jobs will we automate?”, leaders need to ask, “Is our organization structurally able to absorb intelligence at scale?” Answering that requires moving from headcount planning to capability mapping, designing work around the interplay between human strengths, judgment, domain expertise, relationship-building, and machine strengths such as pattern recognition, generation, and orchestration. It means treating architecture as a product: standardizing interfaces, workflows, and data contracts so AI can plug into work without bespoke integration every single time. It means tracking how many workflows, decisions, and customer journeys are genuinely enhanced by AI, not just how many licenses have been bought. And it means steering reduction, augmentation, redesign, and autonomy as one coherent portfolio of change, not four disconnected projects.

Conclusion: The Real Stress Test Is Your Operating Model

AI is already changing jobs. The real test is whether your operating model can evolve quickly enough to harness that change, or whether AI will simply accelerate you toward the limits of the system you already have.

If you would like more information, drop your details in here.

Meike Escherich - Associate Research Director, European Future of Work - IDC

Analyst Profile
Meike Escherich is an associate research director with IDC's European Future of Work practice, based in the UK. In this role, she provides coverage of key technology trends across the Future of Work, specializing in how to enable and foster teamwork in a flexible work environment. Her research looks at how technologies influence workers' skills and behaviors, organizational culture, worker experience and how the workspace itself is enabling the future enterprise.

Key figures at a glance

  • ¥1,304B – IT modernization services market size, 2025
  • 10.2% – Projected average annual growth rate, 2025–2030
  • ¥2,123B – Forecast market size by 2030
  • ~80% – Large and mid-sized enterprises still running legacy systems

Why Japan is outpacing the world

Japan’s IT services market is forecast to grow at a CAGR of 6.6% from 2024 to 2029, nearly double the global average of 3.6%. The answer is structural. Japan carries a uniquely heavy legacy burden, decades of investment in proprietary mainframe environments, complex bespoke systems, and a workforce that has long maintained them. Now, three forces are converging to make modernization unavoidable:

  • Fujitsu Mainframe Sunset – In 2022, Fujitsu announced the end of sales and support for its mainframe and UNIX server products around 2030. This single announcement put more than 1,000 enterprises on an irreversible countdown, accelerating timelines across the entire Japanese market.
  • AI Readiness Imperative – AI adoption presupposes tightly integrated data pipelines and modern business process architectures, exactly what legacy systems make impossible. Modernization is no longer optional for companies that want to remain AI-competitive.
  • Demographic Pressure – The generation of engineers who built and maintained Japan’s legacy systems is retiring. Organizations face a narrowing window to migrate knowledge and infrastructure before institutional memory disappears entirely.

Three paths to modernization

IDC segments IT modernization services into three execution types, each with distinct implications for services firms:

  • Rehost – Lift-and-shift to non-legacy platforms. Preserves existing application assets. The near-term entry point for enterprises constrained by budget or migration timelines.
  • Rewrite – Convert legacy source code to modern languages without changing business logic. A middle path for controlled transformation.
  • Rebuild – Redefine processes, data models, and architecture from the ground up. The highest-value, highest-complexity path.

Near-term, rehost is the second-largest segment after rebuild, driven by enterprises responding urgently to mainframe end-of-life deadlines — though it has already reached maturity and is forecast to decline. The mid-to-long-term growth opportunity lies in application modernization, rewriting, refactoring, and the adoption of microservices and cloud-native architectures.

What enterprises need from services firms

IDC surveyed large and mid-sized Japanese enterprises and found that organizations with significant legacy exposure do not simply want technical execution, they want transformation partners. Security remains a baseline expectation, but top-ranked needs now include business process redesign and cloud architecture strategy.

Demand signals also diverge meaningfully by sector:

  • Financial Services – Prioritizes cloud-native application development capabilities, the ability to innovate rapidly on modern infrastructure.
  • Manufacturing and Distribution – Prioritizes business process transformation, embedding efficiency and intelligence into operations, not just upgrading the underlying technology.

Across all sectors, IDC observes a consistent shift in enterprise expectations: business outcomes are becoming the primary purchase criterion. Technical competence is assumed; value creation is the differentiator.

How to build a winning position now

For services firms, the competitive imperative is clear. The service providers best positioned to win this market will do three things:

1. Codify your legacy modernization track record

Past engagements are an underutilized asset. Service providers should build structured libraries of business outcomes achieved, cost reductions, cycle time improvements, AI readiness unlocked and make these the core of their go-to-market narrative.

2. Develop industry-specific reference architectures for the AI era

Generic modernization pitches are losing traction. Enterprises want system architectures and implementation roadmaps calibrated to their sector, their regulatory environment, and their AI ambitions.

3. Invest in application modernization capabilities ahead of demand

The rehost wave is already approaching its peak. The high-margin opportunity –  rewrite, refactor, rebuild – is building behind it. Service providers who develop deep cloud-native and microservices capabilities now will be the ones enterprises turn to in the second half of this decade.

About the IDC Report

IDC has published a comprehensive analysis of Japan’s IT modernization market: 2026 Japan IT Modernization Market Analysis. The report provides a medium-term market forecast for IT modernization of legacy systems — a primary growth driver in IDC’s Japan IT services market outlook. Legacy systems are characterized by aging and obsolescence, excessive complexity and scale, and a lack of transparency. It covers enterprises’ IT modernization trends and an analysis of the service vendors’ services trend. Market forecasts are segmented by service type, execution type (rehost, rewrite, rebuild), system type, and industry vertical. Together, these analyses offer a comprehensive view of shifting enterprise needs, emerging market opportunities, and the strategies and service offerings of leading vendors in Japan’s IT modernization landscape.

For more detailed insights and market trends, please contact our analysts by completing this form IDC | Identifying Market Opportunities – Contact Us.

Masaru Muramatsu - Senior Research Analyst, Software, Services, and IT Spending, IDC Japan - IDC Japan

Analyst Profile
Masaru Muramatsu is a senior research analyst, responsible for research and analysis of the Japanese IT services market, including IT consulting, systems integration, business services.