AWS Summit 2026: Amazon Quick, Transform, AgentCore Harness, Policy

The big picture from AWS Summit and analyst meetings revolved around a systems approach to AI agents, removing bottlenecks and establishing the company's position as a software vendor with Amazon Quick as the headliner.

Overall, AWS is positioning a portfolio of agentic offerings that work together. Constellation Research analyst Holger Mueller noted:

"AWS is broadly pushing its agentic agenda on both the infrastructure and platform level, with Quick in the apps level. It also is bolstering it's capabilities for SDLC with an always on platform with Continuum. And then there are packaged agents, with the continuous modernization agent easily being the most impactful of all announced agents. Of course AWS has to get it right first."

More AWS Summit 2026:

• QuEra CEO: Quantum moves beyond science project to engineering challenge
• AWS really wants you to try Amazon Quick, eyes enterprise adoption
• AWS sics Transform on technology debt

Here's a look at the news and what you need to know in order of importance.

Amazon Quick

To say AWS management is bullish about Amazon Quick would be a huge understatement. AWS is convinced that Amazon Quick will reinvent how it works as well as its customers. One executive described Quick as his soul mate. He may have been joking, but probably not.

Amazon Quick received a lot of play at AWS Summit and the goal for AWS is to remove what the company calls "the work bottleneck" and the time lost wading through files, folders and apps to get things done.

The company sees Amazon Quick as its unified, context-aware productivity tool that can navigate structured and unstructured data.

Swami Sivasubramanian, VP of AWS Agentic AI, said during is AWS Summit keynote the interactions are driving context and becoming an asset. "More interactions mean more context, and more context means better outcomes, and better outcome means you build more trust, and when you trust it more, you end up handing them more things. This is a virtuous cycle, and that is compounding momentum. Once you have it, the gap between you and everyone grows every single day. This compounding momentum is now one of your most valuable assets," said Sivasubramanian.

Amazon Quick is set up to collect those interactions in a knowledge graph to connect systems. AWS is pushing Amazon Quick as a productivity tool and an abstraction layer for siloed enterprise applications.

Amazon’s Quick Knowledge Graph complements a bevy of application connectors and model context protocol connections. AWS noted customers using Amazon Quick include GoDaddy, the NFL, the NBA and others. Amazon Quick is generally available now and rest assured AWS will have a steady drumbeat of features leading into AWS re:Invent 2026.

The company is getting autonomous agents added to Quick. Quick Autonomous Agents can do the following:

• Run continuously on long-running projects.
• Trigger or schedule desktop events.
• Integrate tools and skills.
• Complete goal-based outcomes autonomously.

AWS execs are using Quick internally for the following:

• Their own daily driver, context‑aware agent for work.
• A cross‑system orchestrator that consolidates tasks and even builds apps from chat.
• A mass‑adoption AI entry point (tens of thousands of users) for both technical and non‑technical workers.
• A strategic lever for using a company’s unique data and context to transform how the business operates.

Transform continuous modernization

Perhaps the most interesting idea from AWS was Transform continuous modernization. Transform is a service designed to migrate legacy infrastructure such as mainframes and .NET to modern code. It's a clunky name, but notable.

Now Transform Zero Debt is aiming to continually analyze code for debt and required migrations and prioritizes them and remediates them over time.

"We're adding continuous modernization, so we transform proactively keep your code bases modern and well documented, so you can unlock engineering capacity and reduce maintenance," said Chief AI and Technology Officer Matt Wood. "Addressing tech debt in waves is like taking out a loan to pay off your credit card. Instead, you can just run this modernization process continuously."

Transform Zero Debt is based on Amazon's internal ASBX (Amazon Software Builder Experience) and could resonate with enterprises that have significant tech debt (in other words all of them).

AWS Transform Zero Debt will surface AWS-driven changes, but also accept other sources of change demands including internal standards, security migrations and platform migrations.

The goal is to earn trust; tie AWS Transform Zero Debt to automated release pipelines and operate continuously.

Continuum

AWS announced Continuum, which brings machine speed to security operations. Think of Continuum as a family of security-focused agents that build on what AWS Security Agent does.

Continuum is an end-to-end security agent that does the following:

• Discover vulnerabilities.
• Proactive threat modeling.
• Validate them with sandboxed dynamic testing.
• Prioritize issues by business impact.
• Remediate.
• And automate ongoing response.

The system starts in learning mode and graduates to enforcement and automated remediation. Wood said Continuum is "a group of agents which act together to use LLMs to drive increases in resolution of security issues."

Chet Kapoor, VP of search, security and observability at AWS, said:

"Continuum works on every vulnerability the way your best engineer works, but not in isolation. It does it against everything it knows about your environment, your architecture, and your business. Then it constructs working code examples in a sandbox environment that provides concrete, reproducible evidence of the issue at each step, continues to generate human reviewable artifacts that are built that will build trust over time."

Agent Core Harness

AWS said it sees AI agents as a model and a harness working together.

The harness handles state persistence, error recovery, context windows, session isolation and orchestration.

AWS announced AgentCore Harness to give enterprises the ability to declare what the agent does, run three simple API calls on isolated microVMs and decouple from models without rewriting agent logic.

Simply put, AWS sees AgentCore Harness doing for AI agents what Lambda did for compute.

AgentCore Harness will also have managed knowledge bases and an agentic retriever that plans queries as well as agentic web search. These features also power Amazon Quick.

Along with the DevOps riff. AWS enhanced its AWS DevOps agent to move from observability to automated release readiness and remediation. Kiro also got play at AWS Summit with a Hiro Mobile app and a DevOps agent that validates infrastructure setup, testing and validation.

AgentCore Policy

AWS outlined agent security and governance tools designed to make it easier to create deterministic guardrails designed to prevent prompt and injection attacks, harmful content and sensitive data exposure.

The company added that it is moving policy for AI agents to Agent Core gateway so agents can't bypass them. AWS outlined how it is moving to session level temporal logic that detects lateral movement and orchestrated attack patterns and enables contingent authorization.

In a nutshell, AWS is looking to build org-level and team-level policies with zero privilege security over time. AWS is looking at building a policy tech stack.