Report and AI Maturity Self-Assessment

State of AI Engineering 2026

Research from Progress and SlashData reveals how 862 software professionals are building, governing and scaling AI - and what separates the most mature organizations from the rest.

Take Your Self-Assessment
  • Less than 5 mins to complete
  • Personalized results & next steps

Developers Expect to Write Less Code. What Replaces it is Still Unclear. 

64%

expect developers to spend less time writing code by hand over the next three years. 

What replaces that work is far less clear.

The research points toward more time spent reviewing AI output, managing security and controlling AI costs.

Developers Expect to Write Less Code. What Replaces it is Still Unclear.

AI is Making Teams Faster. Maintainability is Lagging.

The productivity gains are difficult to ignore. 

But only

72%

report improved delivery speed.

71%

report improved developer productivity.

57%

report improved software quality.

But only

35%

say AI has improved their ability to control technical debt.

AI is accelerating software delivery faster than it is improving teams’ ability to manage what they build. 

Most Organizations are Building With AI. Few Are Operating Systematically.

AI engineering has moved beyond isolated pilots, but systematic production engineering is still the exception. 

Where Does Your Organization Sit?

Benchmark your practices against the research and see where you are strongest - and what may be holding you back. 

FIND YOUR AI ENGINEERING MATURITY STAGE
Most Organizations are Building With AI. Few Are Operating Systematically.

The Safeguards Most Associated with Maturity are Not yet Standard.

6 of 8

production safeguards studied are still relatively underused despite their strong association with higher AI engineering maturity.

Think traceability, logging, clear ownership and the ability to stop AI systems when needed.

The maturity gap is increasingly about operational discipline, not AI access.

The question is no longer just “Can we build it?”

It is increasingly:

Can we understand it, control it and operate it when something goes wrong?

The Safeguards Most Associated with Maturity are Not yet Standard.

Adoption is Ahead of Governance.

Across the five dimensions of the AI Engineering Maturity Index:

Adoption & Integration
scores highest at

0.57

report improved delivery speed.

0.39

The gap matters.

Organizations are getting better at putting AI to work faster than they are building the governance, operational controls and accountability needed to manage it at scale.

The report explores where that gap appears - and what more mature organizations are doing differently.

learn more
Adoption is Ahead of Governance.

As Agents Move from Recommending to Executing, Unintended Actions Rise.

Reported unintended actions:

7%

Agents suggest or recommend

18%

Agents execute within limits

25%

Agents plan and execute objectives

As agents take on more responsibility, engineering controls have to keep pace. 

EXPLORE THE AGENT AUTONOMY FINDINGS
As Agents Move from Recommending to Executing, Unintended Actions Rise.

At the highest autonomy level, reported unintended actions fall to 12%. See the report for analysis.

See Where Your Organization Stands.

Assess your AI Engineering Maturity.
Benchmark your organization across five dimensions of AI engineering maturity and receive:

  • Your maturity stage
  • Your strongest area
  • Your biggest opportunity
  • Peer benchmarks
  • Recommended next steps
prefooter_bg

Additional Resources

Whether you are moving from experimentation into production or optimizing AI systems already operating at scale, these resources can help strengthen the practices behind trustworthy AI engineering. 

Build Trustworthy AI

A practical framework for evaluation, observability, governance and operating AI with confidence. 

Download the Guide
Develop AI Engineering Skills

Explore practical guidance for building and operating AI-enabled applications and development workflows.

Explore AI Engeneering Resources

Progress Solutions for AI Engineering

Progress helps engineering teams build, operate and improve AI-enabled software across the development lifecycle.

Progress
Nexus

Build greater visibility, evaluation and operational control into AI applications and agents as they move from experimentation into production.

Explore Progress Nexus
Progress
Forge

Bring greater structure, context and repeatability to AI-assisted software development and coding-agent workflows.

Explore Progress Forge
AI & Agent-ready Components

Accelerate AI-enabled user experiences with components and patterns designed for modern AI applications.

Explore AI Components
Progress
MCP Servers

Connect AI tools and agents with the development context and capabilities they need through Model Context Protocol integrations.

Explore MCP Servers

AI Adoption is Becoming Common. Engineering Maturity is not.

See where the market is heading — and how your organization compares.

Take Your Self-Assesment
  • Less than 5 mins to complete
  • Personalized results & next steps