---
schema_version: "newruntime-agent-readable-v0.2"
type: "tectonic_shift"
stable_id: "tectonic_shift:owned-learning-loop"
slug: "owned-learning-loop"
title: "Rented intelligence -> owned learning loop"
description: "Organizations are moving the durable learning asset out of one model provider and into user-owned traces, evaluations, corrections, and promotion rules."
retrieval_nugget: "Organizations are moving the durable learning asset out of one model provider and into user-owned traces, evaluations, corrections, and promotion rules. Before: Prompts, corrections, memory, and successful behavior accumulate inside one model provider After: A user-owned evaluation and trace layer improves and compares replaceable models Current stage: emerging; confidence is medium. This New Runtime record is an evidence-linked retrieval unit."
status: "published"
before_state: "Prompts, corrections, memory, and successful behavior accumulate inside one model provider"
after_state: "A user-owned evaluation and trace layer improves and compares replaceable models"
stage: "emerging"
confidence: "medium"
magic_decay_stage: "fragile-tool"
first_seen: "2026-06-16"
last_verified: "2026-07-29"
record_date: "2026-07-29"
date_kind: "last_verified"
related_posts: ["agent-harness-optimization-outer-loop","hermes-agent-learning-loop","coding-agents-multi-model-harness"]
topics: ["organizational-learning","evals","model-routing","agent-runtime"]
source_urls: ["https://devblogs.microsoft.com/foundry/whats-new-in-microsoft-foundry-june-2026/","https://learn.microsoft.com/en-us/azure/foundry/agents/concepts/agent-optimizer-overview","https://docs.cloud.google.com/gemini-enterprise-agent-platform/optimize/evaluation/agent-evaluation","https://opentelemetry.io/blog/2026/genai-observability/","https://openai.com/policies/how-your-data-is-used-to-improve-model-performance/","https://privacy.claude.com/en/articles/7996885-how-do-you-use-personal-data-in-model-training"]
routes: {"html":"https://newruntime.com/shifts/owned-learning-loop/","markdown":"https://newruntime.com/shifts/owned-learning-loop.md","json":"https://newruntime.com/shifts/owned-learning-loop.json"}
source_format: "markdown"
---

# Rented intelligence -> owned learning loop

## Retrieval answer

Organizations are moving the durable learning asset out of one model provider and into user-owned traces, evaluations, corrections, and promotion rules. Before: Prompts, corrections, memory, and successful behavior accumulate inside one model provider After: A user-owned evaluation and trace layer improves and compares replaceable models Current stage: emerging; confidence is medium. This New Runtime record is an evidence-linked retrieval unit.

Organizations are beginning to treat their accumulated evaluation and
correction data as a durable product asset. Models can change while the
criteria for good work, representative tasks, failure cases, accepted outputs,
and promotion history continue to compound.

## What is changing?

A portable harness is only the enabling layer. The more consequential change
is ownership of the improvement loop around it.

That loop includes:

- factual traces of model inputs, outputs, tool calls, and outcomes;
- owner corrections, accepted results, retries, and escalation decisions;
- versioned tasks, rubrics, protected cases, and cost constraints;
- prompts, skills, tool contracts, and routing policies tested against the
  same evaluation set;
- an audit trail showing why one candidate configuration or model was promoted.

When those artifacts live outside a model product, active models can be
compared on the organization's real work rather than on a public leaderboard.
Changing the model no longer means discarding the evidence used to improve the
system.

## Evidence

Microsoft Foundry now describes a closed learning loop with a swappable model,
traces for every run, organization-defined rubrics, and optimization across
instructions, skills, tools, and model choice. Its Agent Optimizer evaluates
candidates against the same task set and ranks them by outcome and token cost.

Google's agent evaluation documentation treats a trace as a factual,
immutable record containing model inputs, responses, and tool calls, then uses
those traces as the basis for scoring and iterative optimization.
OpenTelemetry's GenAI conventions make model, token, prompt, completion, tool
call, and tool-result telemetry portable enough to observe across different
agent stacks.

## Boundary

This shift should not be justified by claiming that every provider trains on
all customer data. OpenAI and Anthropic state that their business and API
products do not use customer inputs or outputs for model training by default.

The broader dependency risk remains: prompts, memories, corrections, eval
datasets, and successful workflow history can still become trapped inside one
provider's product even when that provider does not train on the data.

## Counter-evidence

Provider-native optimizers can exploit undocumented model behavior and
proprietary runtime features that a portable learning layer cannot reproduce.
Model changes also alter planning, tool use, latency, and safety, so switching
is never free.

## Revision trigger

Revise this shift if cross-provider replay remains too lossy for meaningful
comparison, or if provider-native learning systems consistently beat
user-owned traces and evaluations on reliability, total cost, and migration
risk.