📊 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
Most AI ‘agent’ launches in 2026 are actually features layered on existing infrastructure, not independent platforms. This mislabeling affects enterprise security, control, and procurement practices.
Recent industry observations reveal that approximately 90% of AI ‘agent’ launches in 2026 are not true autonomous agents but are instead features built on top of existing vendor infrastructure, misleading enterprise buyers about their capabilities and control.
In May 2026, a vendor announced an AI agent product marketed as transforming knowledge work, priced at $30 per seat per month, with a target of 4,000 paid seats by year-end. However, closer inspection shows that these so-called agents are primarily chat boxes connected to SaaS platforms via OAuth, lacking runtime environments, state persistence, or governance mechanisms.
This pattern exemplifies what industry insiders now call the ‘agent trap’—the widespread practice of labeling feature-like tools as full-fledged agents to inflate perceived value. According to sources, an estimated 90% of such launches are essentially features, not infrastructure, with only about 10% qualifying as genuine platform plays that support portability, governance, and independent operation.
Experts warn that this mislabeling complicates enterprise procurement, as distinguishing between real and fake agents requires new skills. True agents operate autonomously, persist state in customer-controlled stores, and can be swapped or upgraded without losing context, unlike feature implementations tied tightly to vendor infrastructure.
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.
Principles of Agentic AI Governance: A Playbook for Managing AI Risk, Fairness, and Compliance (Agentic Governance and Architecture)
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.
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
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
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.
Impacts on Enterprise Security and Control
This trend matters because enterprises risk increased vendor lock-in, reduced control over their workflows and data, and potential security vulnerabilities. Mislabeling features as agents inflates costs and obscures the true capabilities and risks of AI tools, complicating procurement and governance. Recognizing the difference is essential for making informed decisions that protect organizational assets and ensure operational resilience.Evolution of ‘Agent’ Definitions and Market Trends
Before 2024, ‘agent’ in software referred to processes that ran continuously, maintained state, and were externally governable—characteristics that ensured autonomy and resilience. However, by 2026, the term has been broadly co-opted, often applied to simple chat interfaces or feature layers that lack core agent properties.
The industry’s push to rebrand basic integrations as ‘agents’ is driven by marketing, vendor monetization strategies, and the desire to appear innovative. Major enterprise software providers like Salesforce, ServiceNow, and Microsoft are emphasizing ‘agent platforms,’ but their offerings often resemble headless data models that read and write directly to enterprise systems without true autonomy or portability.
“90% of ‘AI agent’ launches in 2026 are effectively features, not infrastructure. The label is used primarily for marketing and pricing.”
— Thorsten Meyer
Extent and Impact of the ‘Agent Trap’ Phenomenon
While industry estimates suggest that 90% of AI launches are features rather than true agents, precise quantification remains challenging due to inconsistent definitions and proprietary marketing claims. The long-term impact on enterprise security and control is still being assessed as the market evolves.
Emerging Standards and Procurement Practices
Expect increased efforts to develop industry standards for what constitutes a true AI agent, alongside training for procurement teams to better evaluate AI tools. Vendors may face pressure to clarify the capabilities and boundaries of their offerings, while enterprises will need to refine their evaluation criteria to avoid falling into the ‘agent trap.’ The focus will likely shift toward ensuring portability, governance, and control in AI deployments.
Key Questions
What defines a true AI agent in 2026?
A true AI agent operates autonomously, maintains persistent and controllable state, can be swapped or upgraded without losing context, and emits security-compliant audit logs.
Why are so many AI launches labeled as agents if they are not?
Marketing and pricing strategies drive vendors to label feature-layer tools as agents to command higher prices and create a perception of innovation, despite lacking core properties.
How can enterprises avoid buying false ‘agents’?
By applying a five-point filter assessing runtime independence, model swapability, state ownership, auditability, and portability before procurement decisions.
What risks do feature-like ‘agents’ pose to organizations?
They increase vendor lock-in, reduce control over workflows and data, and may introduce security vulnerabilities due to inadequate governance and audit capabilities.
Source: ThorstenMeyerAI.com
