📊 Full opportunity report: The Agent Trap: Why 90% of AI “Launches” Are Infrastructure Liars on ThorstenMeyerAI.com — validation score, market gap, and execution plan.
TL;DR
In 2026, 90% of AI ‘agent’ launches are misrepresented features relying on vendor infrastructure, not independent, governable agents. This impacts enterprise dependency and procurement skills.
Most AI products marketed as ‘agents’ in 2026 are actually features layered on vendor infrastructure, not true autonomous agents, according to recent industry analysis. This mislabeling affects enterprise dependency and procurement strategies, making it a critical issue for organizations adopting AI.
In May 2026, industry analyst Thorsten Meyer highlighted that approximately 90% of AI ‘agent’ launches are merely features on top of vendor-controlled infrastructure. These products typically lack runtime autonomy, state persistence, or governance capabilities, despite being marketed as agents with full autonomy.
Examples include chat boxes that summarize meetings or connect to SaaS platforms via OAuth, but lack the core qualities of a true agent, such as persistent state, external governance, or autonomous operation. The remaining 10% are genuine platform plays, offering infrastructure that allows portability, model swapping, and enterprise control.
This distinction is now a procurement skill, as many enterprises are unknowingly inheriting dependency on vendor infrastructure when they purchase so-called ‘agent’ solutions. The industry has shifted from the traditional definition of an agent—an autonomous, governable process—to a marketing label that inflates product value without delivering true autonomy.
The agent trap.
Why 90% of AI “launches” are infrastructure liars.
A vendor announces an “AI agent.” The product is a chat box that summarises meeting notes — wired to a SaaS via OAuth, no runtime, no audit trail, no portable state. List price: $30 per seat per month. This is the agent trap. The label has been stripped from its meaning. What enterprises are buying — under the word agent — is overwhelmingly a feature on top of someone else’s infrastructure.
Most “agents” are features wearing infrastructure as a costume.
In 2026, the word agent has been stripped from its meaning. Vendors monetize the label. Buyers inherit the dependency. The asymmetry has a number — and the number does the work this story needs.
Applied AI Governance: The Model Context Protocol as an Enterprise Control Plane for Autonomous Agents
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
A request that fails three or more is a feature.
Run the request against five questions before signing any “AI agent” PO. The 90% fail at least three. The 10% pass all five. Price the line item accordingly — because the vendor won’t.
Does it run when no human is logged in?
A real agent runs on a schedule, on a trigger, or as a daemon. If it only works when a user opens a tab, it’s a feature.
Can you swap the model without losing the work?
Real agents treat the model as substitutable. The runbook, tools, memory, and workflow survive a model change. Features are welded to one model.
Where does the state live?
Real agents persist state to a customer-controlled store with a schema you can query. Features persist to “your conversation history” inside the vendor’s database.
What does the audit trail look like to your SOC?
Real agents emit events into a SIEM or webhook stream the security team subscribes to. Features emit nothing — or vendor-side logs you can’t ingest.
What do you keep when the contract ends?
Real agents leave you with skills, prompts, runbooks, memory, integrations as exportable artifacts. Features leave you with the labor you sank into the vendor’s UI — and nothing else.
Practical MLOps: Operationalizing Machine Learning Models
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Salesforce isn’t selling agents. It’s removing the seat.
The dominant 2026 enterprise pattern is “headless 360” — the same Customer 360 / Employee 360 data model the suite sold for two decades, except agents now read and write directly. SDR · CSM · support agent are increasingly configurations of an agent runtime, not job descriptions for human seats.
The 9% genuinely AI-driven layoffs cluster exactly where headless is shipping.
Tier-1 support, junior software engineering, structured-data work — paying customers of a UI. If agents become the operators, the seat license attached to the human disappears. The vendor still gets paid; they just get paid per agent action instead of per human login.
Before · Per-seat humans
After · Headless 360
AI Agents for Beginners: Build Your First Automation Without Coding — The No-Code Guide to n8n, LangChain, and AI Tools That Actually Work
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
A feature cannot be routed.
When you buy a feature agent from a SaaS vendor, you commit to whatever model the vendor chose, at whatever margin the vendor charges. Real infrastructure exposes the model layer. If the vendor can’t tell you what model is running underneath, that is the answer.
QUERY
Platform Engineering for Artificial Intelligence: Designing scalable infrastructure, data pipelines, and model lifecycle management for generative AI and agentic protocols (English Edition)
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
The leverage moves to whoever owns the motherboard — not the chip.
Claude is increasingly the engine inside other people’s products. Legal-tech vendors, customer-success platforms, contract-review startups. This is the Intel Inside playbook. The implication for buyers is not “therefore buy Anthropic.” It is the reverse.
Built on a single closed model.
Brand sits on top of someone else’s chip. Looks like a platform. Priced like one.
- Cabinet vendor sells the platform pricing
- Chip vendor (Anthropic / OpenAI) sets margin
- If the chip vendor moves up the stack, cabinet gets squeezed
- Customer keeps nothing portable when leaving
Runtime that uses models.
Routing, governance, audit, skills layer. The chip is replaceable. The motherboard captures value.
- Multiple models, swappable per-request
- Customer-controlled governance plane
- Skills + integrations are exportable artifacts
- Survives the chip vendor moving up the stack
Skills are the portable infrastructure.
A skill written for Claude Code can be loaded into Codex, into Cursor, into any agent runtime that understands the format. The skill is the IP the customer wrote. The model is the chip. A buyer with 40 skills against an internal runtime can swap the model layer in an afternoon.
declarative · versioned · portable
If the vendor cannot or will not tell you what model is running underneath, that is the answer. You’re not buying an agent platform. You’re buying a wrapper.
Five questions any executive can ask in any vendor pitch.
- Does it run when no human is logged in?
- Can I swap the model without breaking the workflow?
- Where does the state live, and can I query it directly?
- Does it emit events my SOC can ingest?
- When the contract ends, what do I keep?
Four assignments. By role.
Run the five-point filter against every agent line item.
Reclassify each as feature or infrastructure. Re-price accordingly. The exercise will recover budget — usually significant budget.
Inventory the OAuth scopes granted to feature agents.
After Vercel, the agent supply chain is your perimeter. Tokens granted to chat-box agents holding Workspace, GitHub, and CRM scopes are the largest unmanaged risk in the stack.
Per-seat agent SaaS is the most expensive way to buy LLM compute.
Per-action and per-token routing typically costs 60–85% less for the same throughput. Demand the comparison. Vendors that refuse to provide it have answered the question.
Add “AI infrastructure vs feature” to the quarterly risk review.
If management cannot draw the line, the line has not been drawn — and someone else is drawing it for you, on a price tag.
Implications of the ‘Agent’ Mislabeling for Enterprises
This misrepresentation impacts enterprise security, control, and long-term operational flexibility. Organizations relying on feature-based ‘agents’ risk vendor lock-in, reduced control over data and workflows, and potential security vulnerabilities, as these products often lack transparent audit trails and portability. Recognizing the difference is essential for making informed procurement decisions and avoiding dependency on vendor infrastructure that may not meet enterprise governance standards.How the ‘Agent’ Concept Has Evolved by 2026
Historically, an ‘agent’ was a process that ran continuously, maintained state, and was governable externally. Up to 2024, this definition was stable and well-understood. However, in 2026, many products labeled as ‘agents’ are merely chat interfaces or feature layers that call tools or APIs without fulfilling core agent criteria.
The industry has shifted the meaning of ‘agent’ to include any product that appears to perform autonomous tasks, even if it lacks runtime autonomy, persistent state, or external governance. Vendors increasingly market these features as ‘agents’ to command higher prices, despite their limited capabilities.
“90% of ‘AI agent’ launches in 2026 are actually features relying on vendor infrastructure, not true autonomous agents.”
— Thorsten Meyer
What Specific Capabilities Differentiates Real Agents from Features
While the analysis outlines five filters to identify genuine agents, the precise criteria and how they will evolve remain uncertain. Some products may blur lines, and vendors may enhance features to appear more autonomous, complicating clear categorization.
Next Steps for Enterprises and Vendors in AI Agent Adoption
Enterprises should implement the five-point filter before purchasing AI solutions labeled as agents to avoid dependency on vendor infrastructure. Vendors are likely to refine their offerings, but the industry must develop clearer standards and transparency. Future developments may include standardized certification for true agents and increased emphasis on portability, governance, and control features.
Key Questions
How can I tell if an AI product is a true agent or just a feature?
Use the five-point filter: check if it runs without a logged-in user, if the model can be swapped without losing work, where the state is stored, if it provides an audit trail, and what happens when the contract ends. True agents meet all criteria.
Why does it matter if an ‘agent’ is just a feature?
Features rely on vendor infrastructure, creating dependency, reducing control, and increasing security risks. True agents offer portability, governance, and long-term operational independence, which are critical for enterprise security and compliance.
Are there any genuine AI agents available today?
According to industry analysis, about 10% of launches are genuine platform-based agents that support portability, governance, and autonomy. These are still emerging and require careful evaluation.
What risks do enterprises face by mislabeling features as agents?
Risks include vendor lock-in, data control loss, security vulnerabilities, and inability to meet compliance standards. Mislabeling can lead to strategic and operational disadvantages.
Source: ThorstenMeyerAI.com
