{ "totalCount": 10796, "pageSize": 10, "pageNumber": 1, "totalPages": 1080, "posts": [ { "post_type": "post", "post_id": 302272, "permalink": "https:\/\/www.forrester.com\/blogs\/introducing-neocloud-and-neopaas-the-next-frontiers-of-the-ai-native-cloud\/", "title": "Introducing Neocloud And NeoPaaS: The Next Frontiers Of The AI-Native Cloud", "date": "Aug 12, 2026 18:11:32", "excerpt": "Last year, we introduced the concept of the\u00a0AI-native cloud,\u00a0as\u00a0we\u00a0observed\u00a0that the cloud industry was moving beyond commodity infrastructure toward platforms purpose-built for generative AI and agentic AI. We also\u00a0identified\u00a0two emerging paths in this transformation: AI infrastructure cloud platforms (neoclouds) and AI-centric\u00a0neoPaaS.\u00a0 One year later, those two paths have evolved from emerging concepts into distinct market categories, […]", "body": "

Last year, we introduced the concept of the\u00a0AI-native cloud,\u00a0as\u00a0we\u00a0observed\u00a0that the cloud industry was moving beyond commodity infrastructure toward platforms purpose-built for generative AI and agentic AI. We also\u00a0identified\u00a0two emerging paths in this transformation: AI infrastructure cloud platforms (neoclouds) and AI-centric\u00a0neoPaaS.<\/span>\u00a0<\/span><\/p>\n

One year later, those two paths have evolved from emerging concepts into distinct market categories, and we just published two dedicated primer research\u00a0to deep dive into their strategic business values and capability architectures.<\/span>\u00a0<\/span><\/p>\n

Neocloud<\/span><\/a>\u00a0represents\u00a0the infrastructure layer of the AI-native cloud. They are\u00a0AI infrastructure cloud platforms purpose-built to deliver high-performance infrastructure for generative AI, agentic AI, and other AI-powered workloads with a primary focus on GPUs and accelerators. Neocloud providers\u00a0optimize\u00a0compute, storage, networking, scheduling, and AI operations as an integrated. Their goal is to provide predictable access to AI capacity, improve GPU\u00a0utilization, reduce infrastructure bottlenecks, and create clearer economics for AI workloads. As the market matures, neoclouds are evolving from GPU-as-a-service providers into full-stack AI infrastructure platforms that support the complete AI lifecycle across training, inference, retrieval, and live agent execution.<\/span>\u00a0<\/span><\/p>\n

\"\"<\/p>\n

NeoPaaS<\/span><\/span><\/a>\u00a0<\/span>represents<\/span>\u00a0the platform layer of the AI-native cloud.\u00a0<\/span>It is\u00a0<\/span>Kubernetes<\/span>-based, AI-centric platform as a service that makes knowledge management the foundation for agentic AI and unifies application development, modernization, and agentic AI workload management into a governed self-service layer.\u00a0<\/span>NeoPaaS<\/span>\u00a0revives the original promise of PaaS for the AI era. It enables enterprises to transform fragmented data into governed knowledge products, standardize agent lifecycles, provide self-service developer experiences, and apply workload-aware controls for inference, retrieval, and GPU consumption. Rather than forcing every\u00a0<\/span>firm<\/span>\u00a0to assemble its own\u00a0<\/span>agent<\/span>\u00a0stack from open-source components,\u00a0<\/span>neoPaaS<\/span>\u00a0packages\u00a0<\/span>them<\/span>\u00a0into a repeatable platform model<\/span>.<\/span><\/span><\/p>\n

\"\"<\/p>\n

The most important takeaway is that neocloud and\u00a0neoPaaS\u00a0are not competing approaches. They address different layers of the same AI-native cloud architecture. Neocloud provides the AI-optimized execution environment.\u00a0NeoPaaS\u00a0provides the governed platform for building, deploying, and operating AI-native applications and agents. Together, they help enterprises balance performance, cost, governance, sovereignty, and developer productivity as agentic AI adoption accelerates.<\/span>\u00a0<\/span><\/p>\n

The AI-native cloud is no longer a future vision. It is rapidly taking shape through these two emerging categories. Technology leaders evaluating their cloud strategies should now consider not only which cloud to use, but also which neocloud and\u00a0neoPaaS\u00a0capabilities are\u00a0required\u00a0to support AI at scale.\u00a0<\/span>\u00a0<\/span><\/p>\n

If you would like to have strategic guidance to assess neocloud and neoPaaS capabilities, and redesign cloud strategies for enterprise AI, please book an inquiry with me<\/a>\u00a0and Lee Sustar<\/a>\u00a0to discuss.<\/span><\/p>\n", "category": [ { "term_id": 2352, "name": "AI Insights", "slug": "artificial-intelligence-ai", "description": "

The integration of artificial intelligence (AI) is revolutionizing how organizations operate, offering unprecedented opportunities to boost efficiency and drive innovation. Yet, alongside this immense potential comes a layer of complexity that requires deliberate strategy. AI is doing more than just enhancing systems; it\u2019s reshaping how organizations allocate resources, advance capabilities, and achieve growth. Its influence touches every corner of an operating model, challenging leaders to not only capture the power of AI but to create meaningful value with it. The path forward is both exciting and intricate, filled with the promise of transformation and the need for thoughtful navigation. Get the latest AI insights and strategic perspectives from Forrester analysts and experts.<\/p>\r\nDiscover how Forrester supports data, AI, and analytics leaders. <\/a>", "permalink": "https:\/\/www.forrester.com\/blogs\/category\/artificial-intelligence-ai\/" }, { "term_id": 2368, "name": "Architecture & Technology Strategy", "slug": "architecture-technology-strategy", "description": "Your architecture & technology strategy determines whether your business uses technology effectively ... or falls behind competitors. Read Forrester's insights on best practices, trends, and emerging tech in the architecture & technology space.\r\n\r\nDiscover how Forrester supports technology executives. <\/a>", "permalink": "https:\/\/www.forrester.com\/blogs\/category\/architecture-technology-strategy\/" }, { "term_id": 2139, "name": "Cloud Computing Trends", "slug": "cloud-computing", "description": "Cloud computing is revolutionizing IT and enabling digital transformation. Read our insights on why it matters and how to get it right.", "permalink": "https:\/\/www.forrester.com\/blogs\/category\/cloud-computing\/" }, { "term_id": 2208, "name": "public cloud", "slug": "public-cloud", "description": "Use of the public cloud is accelerating, supporting enterprise digital transformation efforts. Read our insights on how firms can leverage the power of the cloud while mitigating security vulnerabilities and other risks.\r\n\r\nDiscover how Forrester supports IT leaders.<\/a>", "permalink": "https:\/\/www.forrester.com\/blogs\/category\/public-cloud\/" } ], "meta_desc": "The next frontier of cloud: neocloud and neoPaaS.", "author": "Charlie Dai", "coauthors": "Lee Sustar" }, { "post_type": "post", "post_id": 302567, "permalink": "https:\/\/www.forrester.com\/blogs\/bringing-crypto-agility-and-pqc-visibility-to-the-network-with-nav\/", "title": "Bringing Crypto-Agility And PQC Visibility To The Network With NAV", "date": "Aug 12, 2026 15:05:09", "excerpt": "The Q4 2025 Forrester Wave\u2122 evaluation for network analysis and visibility (NAV) solutions was the first iteration of the research to evaluate vendors on their post-quantum cryptography (PQC) capabilities, a decision that has become more pronounced in 2026. The rationale was compelling then; it is unavoidable now. Most notably, the US federal government has moved […]", "body": "

The Q4 2025 Forrester Wave\u2122<\/a> evaluation for network analysis and visibility (NAV) solutions<\/a> was the first iteration of the research to evaluate vendors on their post-quantum cryptography (PQC) capabilities, a decision that has become more pronounced in 2026. The rationale was compelling then; it is unavoidable now. Most notably, the US federal government<\/a> has moved PQC migration<\/a> from a long-term discussion to a time-bound activity, effectively putting organizations on the clock.<\/p>\n

Although PQC might not be an apt name, it refers to a class of public key algorithms that are considered unbreakable by a sufficiently sized quantum computer. NIST has standardized several PQC algorithms that are rapidly becoming the global benchmark (though a few nations are advancing their own standards or embracing NIST plus additional options).<\/p>\n

The urgency extends well beyond regulatory compliance. The rise of \u201charvest now, decrypt later\u201d attacks fundamentally changes the risk equation. Additionally, recent quantum computing advances from Google<\/a>, Cloudflare<\/a>, and Microsoft<\/a> have pushed some large technology vendors to set their own migration deadlines to 2029. For sectors such as healthcare research and development, where intellectual property, clinical data, and scientific discoveries may retain material value for decades, the threat is particularly acute.<\/p>\n

In short, quantum risk is no longer theoretical. It is a present-day security and resilience challenge with long-term consequences. Organizations can no longer postpone quantum readiness, as the work toward it should have ideally started yesterday.<\/p>\n

Organizations are facing three hard realities:<\/p>\n

    \n
  1. The assumption.<\/strong> For decades, organizations operated under the belief that encryption was effectively unbreakable within any practical timeframe. Hence, it became a binary checkbox \u2014 at least for the majority of organizations. Consequentially, no one ever considered mechanisms for migrating away from these algorithms, thus making them very \u201cbrittle.\u201d<\/li>\n
  2. The challenge.<\/strong> The path to post-quantum cryptography is far more complex than enterprises realize. Many PQC implementations rely on TLS 1.3, forcing enterprises to first complete long-overdue modernization efforts of upgrading from TLS 1.2. The ability to rotate certificates seamlessly is also a prerequisite. With such prerequisites themselves being big, hard-to-scale steps (and in some cases not feasible), organizations are not even able to consider PQC, much less start their PQC journey.<\/li>\n
  3. The solution.<\/strong> PQC-focused vendor solutions, such as QuSecure, AppViewX, and Keyfactor, help address cryptographic discovery, asset inventories, cryptographic metadata, and (most importantly) crypto-agility.<\/li>\n<\/ol>\n

    NAV Is Critical To PQC Readiness, But The Market Lags<\/strong><\/h3>\n

    Today\u2019s NAV market offers limited visibility into post-quantum cryptography. ExtraHop stands out as one of the few vendors providing meaningful PQC visibility. But the depth of visibility is constrained by the vendor\u2019s support for quantum-resistant algorithms, certificates, and the underlying metadata.<\/p>\n

    NAV solutions can provide visibility into algorithms such as Kyber and Elliptic Curve Diffie-Hellman Ephemeral (ECDHE), along with select certificate metadata, including key type and size, SNI hostname, and certificate fingerprints or thumbprints. While valuable, this level of visibility is still at its surface level, as elements such as the root of the issued certificate, who issued it, the certificate hierarchy, and other attributes are left out.<\/p>\n

    NAV vendors, still playing catch-up on post-quantum cryptography, create a significant visibility gap that enterprises cannot afford to ignore. Until vendors close that gap, organizations will need to compensate the lack of visibility with workarounds. Ingesting cryptographic metadata and related context from dedicated PQC solutions already deployed within the environment into your NAV solution helps bridge this gap to some extent. When combined with deep packet inspection (DPI) and broader network threat detection capabilities that PQC-specific solutions lack, this additional cryptographic context becomes more valuable and actionable.<\/p>\n

    The benefits extend beyond threat detection. Such enhanced visibility within a NAV solution can also strengthen initiatives such as microsegmentation, where policy decisions can often lack cryptographic context. By incorporating PQC-related metadata and visibility from NAV into segmentation strategies, organizations can create trust-based architectures that enforce communication only between systems meeting specific PQC requirements.<\/p>\n

    Most importantly, given that NAV technologies are out of band and do not facilitate any form of native response, this combination with PQC-specific solutions facilitates crypto-agility that extends beyond visibility into action via integrations. NAV solutions can leverage these integrations to deploy reverse proxies and establish tunnels that circumvent the need to upgrade legacy protocols and infrastructure. Other integrations, such as within internal hardware security modules (HSMs), help enterprises build and maintain their cryptographic bill of materials (CBOM). As organizations prepare for a multiyear PQC migration, the winners will be those that can translate cryptographic intelligence into operational security outcomes.<\/p>\n

    Let\u2019s Connect<\/strong><\/h3>\n

    Forrester clients who have questions about this topic or anything related to threat intelligence can book an inquiry or guidance session with me<\/a>.<\/p>\n", "category": [ { "term_id": 51101, "name": "CISO Trends", "slug": "ciso-chief-information-security-officer", "description": "The chief information security officer (CISO) role is growing in importance and remit. Discover the latest trends and analysis for CISOs and information security leaders.", "permalink": "https:\/\/www.forrester.com\/blogs\/category\/ciso-chief-information-security-officer\/" }, { "term_id": 2209, "name": "Cloud Security", "slug": "cloud-security", "description": "As both IT vendors and buyers explore the advantages of cloud-based solutions, they must also evaluate and manage cloud security risks. Read more to get the expert guidance required.", "permalink": "https:\/\/www.forrester.com\/blogs\/category\/cloud-security\/" }, { "term_id": 14963, "name": "Cybersecurity Trends", "slug": "cybersecurity", "description": "Stay up-to-date on the cutting edge of cybersecurity with insights on Zero Trust, vendors, regulations, and other privacy & security topics.", "permalink": "https:\/\/www.forrester.com\/blogs\/category\/cybersecurity\/" }, { "term_id": 2117, "name": "network security", "slug": "network-security", "description": "In an increasingly connected world, excellent network security is a business imperative. Read our insights on how your business can protect its networks.\r\n\r\nDiscover how Forrester supports IT leaders.<\/a>", "permalink": "https:\/\/www.forrester.com\/blogs\/category\/network-security\/" }, { "term_id": 2430, "name": "security architecture", "slug": "security-architecture", "description": "As businesses compete to win and retain customers concerned about the privacy of their data, more firms are learning the value of a robust and effective security architecture. Get benchmarks and technical guidance here.\r\n\r\nLearn more about how Forrester supports IT professionals.<\/a>", "permalink": "https:\/\/www.forrester.com\/blogs\/category\/security-architecture\/" }, { "term_id": 51787, "name": "Security management", "slug": "security-management", "description": "", "permalink": "https:\/\/www.forrester.com\/blogs\/category\/security-management\/" }, { "term_id": 51119, "name": "Security Services", "slug": "security-services", "description": "Get information on the latest security services trends and let Forrester help you protect your organization's digital assets.\r\n\r\nDiscover how Forrester supports IT professionals.<\/a>", "permalink": "https:\/\/www.forrester.com\/blogs\/category\/security-services\/" } ], "author": "Jitin Shabadu", "coauthors": "Sandy Carielli" }, { "post_type": "post", "post_id": 302504, "permalink": "https:\/\/www.forrester.com\/blogs\/the-no-regrets-ai-investment-agenda\/", "title": "The No-Regrets AI Investment Agenda", "date": "Aug 12, 2026 14:47:05", "excerpt": "AI is creating real value, but the biggest opportunities will go to organizations that build the right foundations first. Explore five no-regrets investments that can help technology leaders govern, scale, and realize value from AI regardless of how the technology evolves.", "body": "

    The AI industry continues to make sweeping claims about autonomous agents, self-managing workflows, and enterprises run at machine speed. The reality is more complicated. AI is generating real value, but most of that value remains tightly scoped. Coding productivity is improving. Customer support workflows are becoming more efficient. Information work is accelerating. Yet the enterprise-level gains remain difficult to identify.<\/p>\n

    Many commentators respond by arguing that organizations must adopt radically new operating models. Perhaps. But before redesigning decision rights and reporting structures, it is worth asking a more fundamental question: what exactly are we trying to enable?<\/p>\n

    The current conversation often assumes that autonomy is inherently desirable. I am skeptical. We do not maximize autonomy in human organizations. We do not encourage employees to operate without controls, accountability, or supervision. Why would we expect a different answer for software?<\/p>\n

    Autonomy is a design choice. The responsibility of the technology leader is not to maximize it, but to bound and control it.<\/p>\n

    This observation leads to five questions that every organization deploying AI should be asking. Taken together, they define the foundations of bounded autonomy.<\/p>\n

    \"\"<\/a><\/p>\n

    Identity: Who Or What Is It?<\/strong><\/h2>\n

    Most large enterprises already carry substantial technical debt in digital identity. Over-provisioned service accounts, shared credentials, weak ownership, and unclear accountability are familiar problems. Deterministic software tolerated many of these weaknesses. Goal-seeking systems turn them into active hazards.<\/p>\n

    Recent security research is increasingly focused on agent identity, privilege abuse, and tool misuse. This should surprise no one. An agent can only act through the authority it has been granted. If that authority is poorly governed, the risk follows directly.<\/p>\n

    There is also an economic dimension. Agentic systems create ongoing operational costs, making inventory and accountability prerequisites for effective TokenOps and governance.<\/p>\n

    The initial investment implication is straightforward: know the actors.<\/p>\n

    Organizations will need the equivalent of an application portfolio for agents. Agent identities should be distinct. Their sponsors should be known. Their permissions should be bounded. Every agent should trace back to an accountable human authority. They should also have an explicit lifecycle, including retirement and decommissioning. While not glamorous, inventory is still foundational.<\/p>\n

    See the AEGIS framework<\/a> from our colleagues in Forrester\u2019s Security & Risk service.<\/p>\n

    Capability: What Can It Do?<\/strong><\/h2>\n

    AI capability remains remarkably jagged. A system may perform brilliantly on one task and fail unexpectedly on an adjacent one. We continue to see examples of models achieving extraordinary results on sophisticated benchmarks while struggling with activities that humans find routine. Benchmark performance is useful evidence. It is not operational assurance. The \u201cjagged technological frontier\u201d remains very real.<\/p>\n

    The corresponding investment is to equip the actors.<\/p>\n

    This sounds revolutionary until you look closely. MCP may be new, but APIs are not. Platform engineering and reusable business services are not new. The organizations best positioned for agentic AI are frequently the same organizations that have spent the last decade building internal platforms and treating technology capabilities as products.<\/p>\n

    Lendi provides a useful example. Its AI strategy is built on substantial prior investment in platform services, shared data resources, orchestration capabilities, and reusable business functionality. Their mortgage agents succeed because they stand on top of a platform foundation.<\/p>\n

    One of the genuinely new ideas emerging in AI is the concept of reusable skills. We now have emerging standards for packaging instructions, resources, and code into portable capabilities that can be reused across agents and environments.<\/p>\n

    However, possessing a skill is not the same thing as demonstrating competence. Proficiency is established in the real world, under real constraints, serving real customers, and there is limited real world demand.<\/p>\n

    Meaning: How Does It Understand?<\/strong><\/h2>\n

    Semantic fragmentation is a growing AI hazard. What happens when one agent has the enterprise definition of customer, another inherits a vendor definition, and a third relies on a departmental interpretation? Human organizations have wrestled with these problems for decades. AI amplifies them.<\/p>\n

    A healthcare executive recently described this problem to me as the equivalent of drug interactions. Any individual definition may be fine. The unexpected effects emerge when the definitions interact.<\/p>\n

    The investment implication is to ground the actors with context.<\/p>\n

    Metadata, ontologies, semantic models, knowledge graphs, capability maps, and context graphs all become increasingly important. To be clear, we are looking for semantic alignment<\/strong>, not semantic unification<\/strong>.<\/p>\n

    A common objection is that increasingly capable models will simply infer meaning from messy enterprise environments. Perhaps they will infer meaning more effectively than they do today. The harder problem is authority. Which definition of \u201ccustomer\u201d is the sanctioned one for a regulated process? Which definition governs a financial report? Those are governance questions, not inference questions, and must remain deterministic.<\/p>\n

    What is needed is a dynamic, learning, navigational infrastructure: a way for humans and machines to understand how concepts (which themselves evolve and drift) relate across organizational boundaries.<\/p>\n

    Confidence: How Do We Know It\u2019s Right?<\/strong><\/h2>\n

    Software engineering has long distinguished verification from validation. Building the thing right is different from building the right thing. The distinction matters even more with AI.<\/p>\n

    The investment is assurance through guardrails and evaluation, and again this is not new \u2014 precursors are clear to see in DevOps practices of continuous integration and delivery, policy as code, and the like.<\/p>\n

    One intriguing development is the growing use of AI itself as part of the evaluation process. We are also seeing organizations encode architectural standards, security policies, and development conventions directly into AI working environments. The objective is simple: influence outputs as they are generated rather than auditing them after the fact. Evaluation is the emerging control of AI output using techniques like \u201cLLM as judge.\u201d<\/p>\n

    As accountability requirements increase, so will the investments required to sustain confidence.<\/p>\n

    Control: How Do We Keep It Aligned?<\/strong><\/h2>\n

    Governance ultimately is a problem of feedback.<\/p>\n

    This has been the trajectory of enterprise IT. From agile to continuous integration to continuous delivery to DevOps and product management, all reflect the same underlying idea: faster learning through tighter feedback loops.<\/p>\n

    AI accelerates.<\/p>\n

    The most credible visions of AI autonomy center on feedback loops. The idea is that an AI system can take action, observe the consequences, evaluate the results, and incorporate what it learns into future behavior. Product leaders should recognize this immediately. It is simply the product feedback loop operating at machine speed.<\/p>\n

    Organizations need continuous visibility into agent behavior, outcomes, costs, and risks. They need the ability to intervene, redirect, and recover when systems behave unexpectedly. They need situational awareness rather than periodic inspection. And they need feedback loops capable of operating at the speed of the systems being governed. And systems to manage systems of such loops.<\/p>\n

    Analysts and advisors have been recommending these investment categories well before ChatGPT, and many organizations have been maturing all these dimensions, to their benefit. This is why we consider them \u201cno regrets\u201d investments, with value across a wide range of possible AI futures.<\/p>\n

    Start with inventory and accountable identity. Build governed capabilities on top of that foundation. Ground them in enterprise context. Add assurance proportionate to risk. Finally, create the visibility and feedback loops needed to steer the system in operation.<\/p>\n

    Every one creates value today.<\/p>\n

    The future of AI remains uncertain. The infrastructure required to manage it is considerably less so. Organizations that invest in these foundations will benefit regardless of whether the future arrives as a swarm of autonomous agents or simply a steadily expanding collection of increasingly capable software systems. Invest in the foundations. AI is the new forcing factor, and the consequences of neglect will be harder and harder to ignore.<\/p>\n

    To explore these ideas further, join us at one of Forrester\u2019s upcoming Technology & Innovation<\/a> events in Austin, London, or New York City. You\u2019ll gain practical guidance from Forrester analysts and peers on building the governance, capabilities, and organizational foundations required to scale AI and deliver lasting business value.<\/p>\n", "tags": [ { "term_id": 52002, "name": "TI Central 2026", "slug": "ti-central-2026", "description": "", "permalink": "https:\/\/www.forrester.com\/blogs\/tag\/ti-central-2026\/" }, { "term_id": 52001, "name": "TI East 2026", "slug": "ti-east-2026", "description": "", "permalink": "https:\/\/www.forrester.com\/blogs\/tag\/ti-east-2026\/" } ], "category": [ { "term_id": 2242, "name": "Age of the Customer", "slug": "age-of-the-customer", "description": "", "permalink": "https:\/\/www.forrester.com\/blogs\/category\/age-of-the-customer\/" }, { "term_id": 52023, "name": "Agentic AI", "slug": "agentic-ai", "description": "", "permalink": "https:\/\/www.forrester.com\/blogs\/category\/artificial-intelligence-ai\/agentic-ai\/" }, { "term_id": 2352, "name": "AI Insights", "slug": "artificial-intelligence-ai", "description": "

    The integration of artificial intelligence (AI) is revolutionizing how organizations operate, offering unprecedented opportunities to boost efficiency and drive innovation. Yet, alongside this immense potential comes a layer of complexity that requires deliberate strategy. AI is doing more than just enhancing systems; it\u2019s reshaping how organizations allocate resources, advance capabilities, and achieve growth. Its influence touches every corner of an operating model, challenging leaders to not only capture the power of AI but to create meaningful value with it. The path forward is both exciting and intricate, filled with the promise of transformation and the need for thoughtful navigation. Get the latest AI insights and strategic perspectives from Forrester analysts and experts.<\/p>\r\nDiscover how Forrester supports data, AI, and analytics leaders. <\/a>", "permalink": "https:\/\/www.forrester.com\/blogs\/category\/artificial-intelligence-ai\/" }, { "term_id": 2351, "name": "CIO insights", "slug": "chief-information-officer-cio", "description": "The Chief Information Officer (CIO) plays a crucial and complicated role in today's enterprises. Read insights on how to balance innovation with practicality; the power of big data with customer privacy; and the opportunities of new tech with its accompanying risks.\r\n\r\nDiscover how Forrester supports technology executives. <\/a>", "permalink": "https:\/\/www.forrester.com\/blogs\/category\/chief-information-officer-cio\/" }, { "term_id": 22010, "name": "Information Technology", "slug": "information-technology", "description": "Information technology is the core of today's businesses. Read insights on transforming information technology from a service function to a key driver of business growth.\r\n\r\nDiscover how Forrester supports IT leaders.<\/a>", "permalink": "https:\/\/www.forrester.com\/blogs\/category\/information-technology\/" }, { "term_id": 2190, "name": "infrastructure & operations", "slug": "infrastructure-operations", "description": "Infrastructure & operations are the foundation on which business success is built. Read insights on keeping infrastructure & operations working efficiently and implementing the latest technologies.\r\n\r\nDiscover how Forrester supports IT leaders.<\/a>", "permalink": "https:\/\/www.forrester.com\/blogs\/category\/infrastructure-operations\/" } ], "meta_desc": "No-regrets AI investments can help technology leaders govern, scale, and create value from AI regardless of how the technology evolves.", "meta_keywords": "no-regrets AI investments, AI investment strategy, enterprise AI strategy, AI governance, agentic AI, AI readiness, scaling AI, AI transformation, bounded autonomy, AI risk management", "author": "Charles Betz", "coauthors": "Manuel Geitz" }, { "post_type": "post", "post_id": 302416, "permalink": "https:\/\/www.forrester.com\/blogs\/anthropics-pricing-shift-puts-ai-consumption-risk-back-on-customers\/", "title": "Anthropic\u2019s Pricing Shift Puts AI Consumption Risk Back On Customers", "date": "Aug 12, 2026 11:43:14", "excerpt": "Back in May of this year, Anthropic announced changes to its pricing model. Its original fixed-fee, per-seat subscription model was replaced with one that separates platform access from AI consumption. Customers still pay for access, but usage is now metered and billed separately based on token consumption. Under the previous model, customers were split into […]", "body": "

    Back in May of this year, Anthropic announced <\/span>changes to its pricing model<\/span><\/a>. Its original fixed-fee, per-seat subscription model was replaced with one that separates platform access from AI consumption. Customers still pay for access, but usage is now metered and billed separately based on token consumption.<\/span><\/p>\n

    Under the previous model, customers were split into two user tiers: premium ($200 per user per month) and standard ($40 per user per month), with API discounts typically ranging from 10% to 15%. Those economics only worked if most users consumed a small fraction of the value included in their subscriptions. Agents broke that assumption. SDKs, automation, GitHub Actions, and scripts can consume at a scale that humans simply are not capable. Anthropic’s new pricing model reflects this shift. Users are now priced by role, with Claude.ai for nontechnical users at $10 per user per month and Claude Code for technical users at $20 per user per month and no API discounts.<\/span><\/p>\n

    The change also extends beyond chat. Services such as the Claude Agent SDK, Claude Code GitHub Actions, the\u00a0claude\u00a0-p command, and third-party applications authenticated through Anthropic consume separate monthly Agent SDK credits. Those credits expire each month and do not roll over. Once exhausted, usage either shifts to standard API pricing or stops until credits reset.<\/span>\u00a0<\/span><\/p>\n

    While Anthropic positioned the changes as a benefit through lower seat prices and added credits, the\u00a0reality\u00a0is costs are now driven by consumption. Token usage, model selection, Claude Code adoption, and commitment levels\u00a0will\u00a0have a far greater impact on spend than seat counts.\u00a0So\u00a0a\u00a0small group of power users can drive a disproportionate share of costs, turning successful AI adoption into both a productivity win and a budgeting challenge.<\/span>\u00a0<\/span><\/p>\n

    Our clients\u00a0have reacted negatively, arguing that tighter usage limits and consumption-based pricing\u00a0massively reduce the\u00a0value that made Claude Code attractive\u00a0to developers. Many have also criticized\u00a0Anthropic’s\u00a0communication\u00a0as unclear about what was changing and why.\u00a0Some believe Anthropic framed the changes as a community benefit when the practical outcome for heavy users will be significantly\u00a0higher\u00a0costs.<\/span>\u00a0<\/span><\/p>\n

    So\u00a0what does this mean\u00a0for\u00a0you? FinOps teams now have another cost layer to manage as consumption-based pricing introduces greater forecasting complexity and budget volatility.<\/span>\u00a0<\/span><\/p>\n

    If this sounds familiar, that’s because Anthropic is heading in the same direction as the rest of the enterprise AI market. I\u2019m seeing a growing separation between platform access and AI consumption.\u00a0<\/span>OpenAI<\/span><\/a>\u00a0combines subscriptions with\u00a0credit- and\u00a0token-based usage.\u00a0<\/span>Google<\/span><\/a>\u00a0measures Gemini AI usage by the compute\u00a0required, not by the request number.\u00a0<\/span>Cohere<\/span><\/a> has long favored consumption-based pricing tied to usage. Although these are new changes, the fixed-fee per user pricing structure was never meant to stay. Rather, it was the \u201cthe first one\u2019s free\u201d approach to entice tech users with a later switch to consumption pricing models. The market is converging on a model in which customers pay for access and pay again for what they consume.<\/span>\u00a0<\/span><\/p>\n

    What customers need to remember is that the most important number in the agreement is no longer the seat price. Instead, it is the usage assumption and forecast used to calculate the commitment.\u00a0<\/span>\u00a0<\/span><\/p>\n

    Remember these three things:<\/span>\u00a0<\/span><\/p>\n