Amazon Unleashes AI Agents: Kiro Codes for Days

AWS introduces three new AI agents, including an autonomous coding assistant named Kiro.

Amazon Web Services (AWS) has unveiled three new AI agents designed to automate complex development tasks. Among them is 'Kiro autonomous agent,' capable of coding independently for extended periods, promising significant changes for software teams.

Katie Rowan

By Katie Rowan

December 10, 2025

3 min read

Amazon Unleashes AI Agents: Kiro Codes for Days

Key Facts

  • Amazon Web Services (AWS) previewed three new AI agents for software development.
  • The 'Kiro autonomous agent' can work independently on coding tasks for days.
  • Kiro learns team workflows and coding standards over time.
  • AWS Security Agent identifies security problems and suggests fixes.
  • DevOps Agent automatically tests code for performance and compatibility.

Why You Care

Ever wish your toughest coding tasks could just… handle themselves? What if an AI could write and debug code for days without constant supervision? Amazon Web Services (AWS) recently previewed three new AI agents, and one, called Kiro, promises exactly that. This creation could fundamentally change how you approach software creation, freeing up your valuable time for more creative work.

What Actually Happened

Amazon Web Services (AWS) recently showcased three new artificial intelligence agents, according to the announcement. These agents are designed to automate various aspects of the software creation lifecycle. They handle tasks like writing code, performing security reviews, and streamlining DevOps processes. Preview versions of these agents are available now for developers to explore. The most notable among them is the “Kiro autonomous agent,” a frontier agent that AWS claims can operate independently for days at a time. This agent builds upon AWS’s existing Kiro AI coding tool, enhancing its capabilities significantly.

Why This Matters to You

Imagine offloading complex, multi-day coding projects to an AI. The Kiro autonomous agent learns your team’s workflow and coding standards, as detailed in the blog post. It then independently tackles assigned tasks from the backlog. This means less manual oversight and more time for your team to innovate. For example, consider a essential code update affecting 15 different corporate software applications. Instead of assigning and verifying each update individually, you could assign Kiro to fix all 15 with a single prompt. How much more productive could your team be with this kind of assistance?

Here’s a breakdown of the new AWS AI agents:

Agent NamePrimary FunctionKey Benefit for You
Kiro autonomous agentIndependent code generation and task completionReduces manual coding effort, handles complex projects
AWS Security AgentIdentifies security issues, suggests fixesEnhances code security, automates vulnerability checks
DevOps AgentTests code for performance and compatibilityEnsures smooth deployment, prevents incidents

AWS CEO Matt Garman highlighted the agent’s learning capability. “It actually learns how you like to work, and it continues to deepen its understanding of your code and your products and the standards that your team follows over time,” he said. This continuous learning ensures the agent adapts to your specific needs.

The Surprising Finding

Here’s the twist: The Kiro autonomous agent maintains “persistent context across sessions,” as mentioned in the release. This means it doesn’t suffer from memory loss, forgetting previous instructions or progress. Unlike many AI tools that require constant re-prompting, Kiro can be given a task and work on it for hours or even days. This is surprising because it challenges the common assumption that AI agents need frequent human intervention. Most current AI assistants struggle with long-term memory and complex, multi-step projects. Kiro’s ability to retain context allows it to handle much larger and more intricate creation challenges independently. This continuous memory is a significant leap forward for autonomous AI agents.

What Happens Next

We can expect to see these AWS AI agents integrated into creation workflows over the next 12-18 months. Early adopters will likely experiment with Kiro for routine but time-consuming tasks. For instance, a small startup might use Kiro to refactor legacy code, freeing up their senior developers. The industry implications are vast, potentially accelerating creation cycles and reducing human error. Developers should start exploring these preview versions to understand their capabilities. Think about how these tools could streamline your current projects. The team revealed that the AWS Security Agent will independently identify security problems as code is written. Meanwhile, the DevOps Agent will automatically test new code for performance and compatibility. These tools aim to create a more automated and efficient creation pipeline for everyone.

Ready to start creating?

Create Voiceover

Transcribe Speech

Create Dialogues

Create Visuals

Clone a Voice