Data-driven analysis of AI API tool adoption patterns, agentic AI growth trajectories, and the infrastructure demands shaping how developers build action-oriented AI systems
The shift from conversational AI to action-oriented agents represents a fundamental change in how businesses deploy artificial intelligence. With 7.53 million AI API calls recorded in the past 12 months—a 40% year-over-year increase—developers are building systems that don't just respond but execute tasks across email, calendars, CRMs, and databases. Arcade.dev sits at the center of this transformation as the MCP runtime that enables and governs multi-user authorization—so agents can act on behalf of real users across a tool catalog of hundreds of enterprise platforms with fine-grained, delegated user permissions and scoped access. Arcade focuses on token and secret management (not handling customer data), making it far easier to operationalize safe tool actions than building and maintaining this layer in-house.
Key Takeaways
- AI API market set to explode 9x – The sector grows from $41.05 billion in 2024 to $373.38 billion by 2032 at 31.79% CAGR
- Developer adoption hits critical mass – 89% of developers now use generative AI in their daily work
- Agentic AI becomes mainstream – By 2028, 33% of enterprise software will include agentic AI capabilities, up from less than 1% in 2024
- Security concerns dominate – 51% of developers cite unauthorized API calls from AI agents as their top security worry
- API revenue generation proven – 65% of organizations now generate revenue from their APIs
- Enterprise AI spending surges – Model API spending more than doubled in under a year, surging from an estimated $3.5 billion in late 2024 to $8.4 billion by mid-2025
- ROI validated at scale – Companies report 3.7x return for every dollar invested in generative AI
The Rise of API Tools as AI's 'Hands': Moving Beyond Conversational AI
1. AI API market valued at $41.05 billion in 2024, projected to reach $373.38 billion by 2032
The AI API sector represents one of technology's fastest-growing markets, with the industry expected to expand from $41.05 billion to $373.38 billion over the next seven years. This 9x growth reflects the fundamental shift from AI as a conversation partner to AI as an action executor. Organizations building agentic systems require robust API infrastructure to connect language models with real-world services.
2. AI orchestration market reaches $11.47 billion in 2025
The market for coordinating AI workflows and tool calls hit $11.47 billion in 2025, up from $9.33 billion in 2024. This growth reflects demand for platforms that manage complex multi-step agent operations. Arcade's architecture addresses this need by handling multi-user authorization, state management, and resilience across distributed tool executions.
3. 89% of developers use generative AI in their daily work
Developer adoption of AI tools reached 89% in 2025, making AI-assisted development the norm rather than the exception. This near-universal adoption creates demand for robust tool-calling infrastructure that can handle the volume and complexity of modern AI applications.
4. Only 24% of developers actively design APIs with AI agents in mind
Despite widespread AI usage, just 24% of developers design APIs specifically for AI agent consumption. This gap creates opportunities for platforms that bridge existing APIs with agent-friendly interfaces. Arcade solves this with a growing tool catalog that makes common services agent-ready while enforcing multi-user authorization and scoped permissions at runtime.
Arcade also provides an MCP framework to build tools—so a tool does not have to be in the catalog to be governed with the same multi-user authorization controls.
5. AI platform market grows from $18.22 billion to $94.31 billion by 2030
The broader AI platform market will expand from $18.22 billion in 2025 to $94.31 billion by 2030, representing a 38.9% compound annual growth rate. This trajectory indicates sustained enterprise investment in AI infrastructure for years to come.
6. Global AI market reaches $391 billion in 2025
The total addressable market for AI technologies hit $391 billion in 2025, establishing AI as a major economic force. Within this massive market, tool-calling and API integration represent essential infrastructure components that enable practical AI deployments.
Securing the Surge: Multi-user Authorization and Governed Tool Use
7. 51% of developers cite unauthorized API calls as their top AI security concern
More than half of developers (51%) identify unauthorized or excessive API calls from AI agents as their primary security worry. This concern underscores the need for multi-user authorization that governs what actions agents can perform once “logged in”—including exactly which permissions and scopes are delegated per user. Arcade addresses this by enforcing fine-grained, delegated user permissions and scoped access at the moment a tool is invoked, backed by centralized token and secret management.
8. 49% worry about AI systems accessing sensitive data they shouldn't see
Nearly half of developers (49%) express concern about AI accessing inappropriate data. This statistic highlights why zero-token-exposure architectures matter for production deployments. Arcade's design ensures tokens remain encrypted at rest and are never exposed to the language models themselves.
9. 36% of developers lack trust in AI systems
Trust remains a significant barrier, with 36% of developers expressing distrust in AI systems. Building trust requires transparent multi-user authorization, audit trails, and explicit scope boundaries—all core features of enterprise-grade tool-calling platforms.
10. 33% have ethical, legal, or compliance concerns about AI
One-third of developers (33%) cite regulatory and compliance issues as barriers to AI adoption. Enterprise deployments require platforms that meet security certifications and provide audit capabilities. Enterprise deployments require tool actions that are fully auditable, scope-bounded, and consistently governed across systems—not stitched together with one-off permission logic.
11. 93% of API teams face collaboration blockers
Collaboration challenges affect 93% of API teams, with inconsistent documentation being a primary issue. This fragmented, domain specific reality makes securing AI-API interactions even more complex, as teams struggle to maintain consistent multi-user authorization patterns across services.
Developer Experience: Accelerating AI Agent Deployment with Pre-built Connectors
12. 7.53 million AI API calls recorded in 12 months, a 40% year-over-year increase
API usage data shows 7.53 million calls to AI APIs in the past year, reflecting explosive growth in AI-powered applications. This volume increase demands infrastructure that can handle scale while maintaining security and reliability.
13. 82% of organizations have adopted an API-first approach
Most organizations (82%) now prioritize API-first development, with 25% operating as fully API-first companies. This architectural shift creates the foundation for AI agent integration, as more services expose programmatic interfaces that agents can use.
14. 69% of developers spend 10+ hours per week on API-related tasks
Developers dedicate substantial time (69% spend 10+ hours weekly) to API work. Platforms that reduce integration complexity directly impact developer productivity. Arcade reduces the hidden effort of building safe tool use by standardizing multi-user authorization and token/secret management across tools—so teams don’t have to reinvent scope mapping, approvals, and auditing for every integration.
15. 68% use AI to improve code quality
The majority of developers (68%) leverage AI for code improvement, demonstrating practical value in development workflows. This adoption pattern extends to AI-assisted API integration and testing.
16. 41% use AI to generate API documentation
Documentation automation reaches 41% adoption, showing AI's impact on developer experience. Clear documentation is essential for AI agents that need to understand API capabilities and parameters.
17. GitHub Actions leads CI/CD adoption at 54%
Development infrastructure preferences show GitHub Actions at 54% market share, ahead of AWS DevOps (34%) and Azure DevOps (29%). This concentration suggests integration patterns for AI tool deployment pipelines.
Unlocking New Horizons: Agentic Commerce and Transactional AI
18. By 2028, 33% of enterprise software will include agentic AI
Gartner projects that 33% of enterprise software will incorporate agentic AI capabilities by 2028, up from less than 1% in 2024. This 33x increase represents one of the fastest technology adoption curves in enterprise software history.
19. 15% of day-to-day work decisions will be made autonomously by 2028
Beyond simple task execution, 15% of daily decisions will be handled autonomously by AI agents by 2028, up from 0% in 2024. This projection emphasizes the need for secure, auditable decision-making infrastructure.
20. AI agents market projected to reach $52.6 billion by 2030
The dedicated AI agents market will grow to $52.6 billion by 2030 at a 45% compound annual growth rate. This market encompasses agents that perform real transactions, not just conversations—making secure payment authorization and transaction controls essential. Arcade's Agentic Commerce suite addresses this with single-use virtual cards and OAuth-style payment flows.
21. 62% of organizations are experimenting with AI agents
Most organizations (62%) have begun AI agent experiments, though production deployments remain limited. This experimentation phase creates demand for platforms that can support proof-of-concept projects while scaling to production.
22. Less than 10% of organizations have scaled AI agents in any function
Despite widespread experimentation, fewer than 10% have successfully scaled AI agent deployments. This gap between experimentation and production often stems from multi-user authorization complexity and security concerns that specialized platforms help address.
Monitoring and Evaluation: Ensuring Reliable AI-Tool Performance
23. 70% of developers are aware of Model Context Protocol (MCP), but only 10% use it regularly
Awareness of MCP reaches 70%, yet only 10% use it in production. This adoption gap reflects challenges in implementing MCP with proper multi-user authorization. Arcade closes this gap by acting as the MCP runtime that enforces multi-user authorization—delegated scopes, tool-level permissions, and auditable actions—when MCP-based agents invoke real tools.
24. AI platform software market reaches $153 billion by 2028
IDC projects the AI platform software market will hit $153.0 billion by 2028 at a 40.6% CAGR. This growth funds continued platform improvements including better evaluation and monitoring capabilities.
25. AI orchestration market reaches $42.3 billion by 2033
Long-term projections show the orchestration market reaching $42.3 billion by 2033 at a 19.28% CAGR. This sustained growth indicates ongoing investment in coordination layers that manage complex agent workflows. Arcade's evaluation framework helps teams automate and benchmark LLM-tool interactions for reliable performance.
Scaling Governed Tool Access Across the Enterprise
26. 78% of organizations use AI in at least one business function
Enterprise AI adoption reached 78% in 2025, up from 55% in 2023. This 23 percentage point increase in two years demonstrates accelerating enterprise confidence in AI deployments.
27. 71% regularly use generative AI in at least one business function
Beyond experimental use, 71% of organizations report regular generative AI usage in production, up from 65% in early 2024. This production adoption requires consistent, governed multi-user authorization across tools so teams can scale without over-permissioning or brittle one-off integrations.
28. Average company implements AI in three different business areas
Organizations deploying AI do so across three different functions on average. This multi-domain usage requires platforms that can support diverse integrations—from Gmail and calendar to Slack messaging to CRM systems.
29. 88% of enterprises report regular AI use
Enterprise adoption reaches 88% globally with regular AI use in at least one business function. This high adoption rate drives demand for enterprise-grade infrastructure with compliance certifications and flexible deployment options.
30. 95% of U.S. companies now use generative AI
Near-universal adoption is reached with 95% of U.S. companies using generative AI in some capacity. This saturation shifts competition from adoption to effective implementation and scaling.
The Power of Ecosystems: Integrations with Leading AI Frameworks and Tools
31. Model API spending more than doubled in under a year
Investment in model APIs grew from an estimated $3.5 billion in late 2024 to $8.4 billion by mid-2025, demonstrating willingness to pay for AI capabilities. This spending increase funds ecosystem development including tool-calling infrastructure.
32. Llama API calls grew 6.9x year-over-year
Open-source model adoption explodes with Llama seeing 6.9x growth in API calls. This diversification beyond proprietary models creates opportunities for tool-calling platforms that support multiple model providers. Arcade integrates with LangChain, while keeping its core value focused on governed multi-user authorization across tools.
Many teams orchestrate agent workflows with LangGraph—LangChain’s graph-based framework for building stateful, multi-step agent flows—then rely on Arcade as the MCP runtime to enforce multi-user authorization and scoped permissions when the workflow takes real tool actions.
33. Gemini API calls grew 3.1x year-over-year
Google's AI APIs showed 3.1x year-over-year growth, indicating strong enterprise adoption of multi-cloud AI strategies. Platforms that remain model-agnostic benefit from this diversification trend.
Community and Support: Scaling AI Tool Adoption
34. AI tools now reach 378 million people worldwide
Global AI tool usage hit 378 million users in 2025, representing 64 million new users added since 2024. This rapid user growth creates scaling demands for underlying infrastructure.
35. 65% of organizations generate revenue from their APIs
API monetization is widespread, with 65% of organizations generating revenue from their APIs. This revenue dependency makes reliable API infrastructure a business-critical investment.
36. Among API revenue generators, 25% derive over half their revenue from APIs
For organizations that generate revenue from APIs, a quarter derive more than half of their total revenue from API programs. This revenue concentration demands enterprise-grade reliability and support. Arcade's pricing structure scales from free tiers for experimentation to enterprise plans with dedicated support.
37. 43% of fully API-first organizations generate over 25% of revenue from APIs
Companies that prioritize API-first development see 43% generating significant revenue from their API programs. This correlation validates investment in API infrastructure and tooling.
38. Companies report 3.7x ROI for AI investments
Return on AI investment reaches 3.7x for every dollar invested, providing clear business justification for expansion. This proven ROI accelerates adoption cycles and infrastructure spending.
Industry-Specific Adoption Patterns
39. 34% of insurers fully adopted AI in 2025, up from 8% in 2024
Insurance sector adoption jumped from 8% to 34%—a 325% increase in one year. This sector leads in transactional AI where agents process claims, assess risk, and manage customer interactions.
40. 77% of manufacturers now use AI solutions
Manufacturing AI adoption reached 77% in 2025, up from 70% in 2024. This 7% increase reflects integration of AI into operational processes requiring tool access to production systems.
41. IT & telecommunications reached 38% AI adoption with $4.7 trillion projected value-add by 2035
The technology sector achieved 38% adoption with projections of $4.7 trillion value creation by 2035. This sector leads in building AI-powered internal tools that require robust API infrastructure.
42. U.S. private AI investment reached $109.1 billion in 2024
American AI investment hit $109.1 billion, nearly 12 times China's $9.3 billion. This investment concentration funds rapid infrastructure development and platform maturation.
43. Generative AI attracted $33.9 billion globally in 2024
Private investment in generative AI specifically reached $33.9 billion, an 18.7% increase from 2023. This focused investment funds continued capability expansion and infrastructure development.
Implementation Best Practices for AI Tool Adoption
Successful AI tool integration requires multi-user authorization, clear scope boundaries, and enterprise governance from day one. The biggest blocker is not “connecting APIs,” but controlling which users an agent is allowed to represent and what exact actions it is permitted to take once connected.
Leaders should implement one high-value use case end-to-end first, get it to production, then scale to additional workflows once multi-user authorization and scope controls are proven.
Organizations scaling from experimentation to production face common challenges that proper infrastructure addresses:
Multi-user authorization and governance:
- Define which users an agent may represent, and enforce least-privilege scopes per tool action
- Centralize token and secret management so credentials don’t sprawl across teams and services
- Maintain complete audit trails for every agent action (who, what tool, what scope, what happened)
Operational controls:
- Put guardrails on rate/cost to prevent runaway tool calls
- Establish approvals for sensitive actions (payments, data exports, account changes)
Why this is hard without Arcade:
- Building this layer internally means continuously mapping scopes across tools, preventing over-permissioning, and maintaining auditing as integrations change—work that becomes brittle and expensive as the tool surface area grows
Arcade addresses these concerns as the MCP runtime for multi-user authorization, with centralized token and secret management and governed, auditable tool actions that scale from an initial production workflow to broader enterprise adoption.
Future Growth Projections
The data points toward sustained, accelerating growth in AI tool adoption. With 78% of organizations already using AI and the market heading toward $373.38 billion by 2032, infrastructure investments made today will compound over the coming decade.
Key growth vectors to monitor:
- Agentic AI expansion: From <1% to 33% of enterprise software by 2028
- Autonomous decision-making: 15% of daily decisions made by AI by 2028
- AI agents market: $52.6 billion by 2030 at 45% CAGR
- Model API spending: Continued doubling as capabilities expand
- Enterprise orchestration: $42.3 billion by 2033 for workflow coordination
Organizations that establish solid AI infrastructure now position themselves to capture value as these trends materialize.
Frequently Asked Questions
What drives the growth in AI API tool usage?
Three factors dominate: the shift from chat-based AI to action-oriented agents that perform real tasks, enterprise demand for AI that integrates with existing business systems, and improving model capabilities that make complex multi-step operations reliable. The 40% year-over-year increase in AI API calls reflects this convergence of capability and demand.
What security measures are essential for AI tool implementations?
With 51% of developers citing unauthorized API calls as their top concern, essential measures include multi-user authorization with explicit delegated scopes and tool-level permissions, token encryption at rest, audit trails for all agent actions, and rate limiting to prevent runaway consumption. Platforms should never expose raw credentials to language models.
What percentage of enterprises have successfully scaled AI agents?
Despite 62% of organizations experimenting with AI agents, fewer than 10% have scaled to production deployments. This gap often stems from multi-user authorization complexity and security concerns that specialized tool-calling platforms help resolve.
What ROI can organizations expect from AI tool investments?
Companies report an average 3.7x return for every dollar invested in generative AI. Organizations generating revenue from APIs see even higher returns, with a quarter of revenue-generating organizations deriving over half their total revenue from API programs.
