As autonomous DevSecOps tools advance rapidly across the industry and major competitors appear vulnerable, AWS this week doubled down on its pitch to enterprise developers with a preview of an update to its AI agents.
At AWS Summit New York on June 17, AWS introduced agent AI updates that include:
AWS Continuum (Preview): Penetration testing and code review launched as AWS Security Agent at re:Invent 2025 adds threat modeling and broader vulnerability management feature set.
AWS DevOps agent release management (Preview): Extend the existing root cause analysis available in DevOps agents to automatically test and validate your code and detect issues before they reach production.
AWS Continuous Modernization Transformation (Preview): Autonomously discover and fix technical debt in your existing codebase, including framework upgrades, dependency patches, and version updates.
Amazon Bedrock AgentCore updates (Preview):Incorporates a new model-agnostic harness for agent deployment. Managed AgentCore Web Search. A managed knowledge base of unstructured data. Enforce enterprise policies and improve the quality of agent output with gateway guardrail integration and agent optimization.
AWS context (Preview):Agent context knowledge graph integrated with Glue Data Catalog, SageMaker Unified Studio, and third-party data sources using APIs.
AWS Kiro app for iOS (Preview): Features an always-on, cloud-based coding agent session so developers never have to interrupt their work.
Amazon Quick autonomous agent: Amazon, joining the Claw boom in autonomous personal desktop agents, includes agents that automatically update business context and can run without human intervention. New MCP connectors to code-free agent builder and third-party apps.
Rapid evolution of Agentic DevSecOps, intense competition
Dr. Matt Wood, chief AI and technology officer at AWS, said in a press conference this week that AI agents have evolved across industries from assistants to autonomous actors in just a few months.
“Over the past three to four months, we have seen a step-function increase in the cybersecurity capabilities of our large language models, resulting in a significant expansion of the security capabilities available to our customers today…under a single umbrella called AWS Continuum,” Wood said.
Chet Kapoor, Vice President of Security Services and Observability at AWS, introduces new AWS Continuum services at AWS Summit New York.
AWS Continuum shifts the focus of security agents from detecting security vulnerabilities to solving them on their own using what they know about the business context of the application code, Chet Kapoor, vice president of security services and observability at AWS, said in the summit keynote.
“Continuum investigates every vulnerability the way a good engineer would, but not in isolation, but against everything we know about your environment, architecture, and business,” said Kapoor. “We then build working samples in a sandbox environment that provide concrete, reproducible evidence of the problem.”
Kapur said that ongoing remediation and mitigation in production environments will be verified using the same process.
IDC analyst Katie Norton said this kind of functionality is becoming a key element among enterprise vendors, noting that numerous competitors have introduced similar capabilities, including Google Cloud, Microsoft’s GitHub, GitLab, Harness, and Datadog.
“The movement from detection to resolution (scan, verify, fix, verify) is now the common direction across the field,” Norton wrote in an email to TechTarget this week. “We are probably in the early adopter/early majority stage across vendors in terms of offering this as an agent feature.”
However, another new AWS Continuum feature, automating threat modeling during the design phase, “seems fresher,” in Norton’s opinion.
“AWS had some ‘old school’ threat modeling capabilities, but the market is moving more towards AI here. [specialists] “Companies like Apiiro, Prime Security, DevArmor, Clover, and a few others doing continuous threat modeling.” “It’s good validation to see larger companies offering comparable capabilities,” she writes.
AWS puts developers in a bind with GitHub
These updates come as Microsoft GitHub, which used Copilot to dominate in AI coding assistants and agents, has faced reliability issues and developer concerns over usage-based billing changes over the past six months. According to Moor Insights & Strategy analyst Jason Andersen, this leaves GitHub open to disruption by competitors.
AWS products are better, but GitHub’s prices are a little worse.
jason andersen Analyst Moore Insights and Strategies
“The AWS story for developers is getting a lot better…If you look at how Kiro works with some of the new AgentCore services like Harness and Policy, it includes more of a lifecycle management perspective. For me, that was the big advantage of GitHub over agents,” Andersen said. “If you look at GitHub, the new cost model is a challenge. Frankly, GitHub Copilot was a great value, but now it’s more in line with the rest of the market. So AWS products are better, and GitHub’s pricing is a little worse.”
“For organizations that have already invested in GitHub and Microsoft Azure, the switch is much more difficult and requires an extensive comparison of the DevSecOps capabilities of the underlying cloud platform and its agents,” said Torsten Volk, an analyst at Omdia, a division of Informa TechTarget.
“GitHub is part of the Microsoft stack as a whole, and if you no longer like the value backing, you’re generally more likely to move to AWS or GCP,” he said. “But leaving GitHub behind is also a huge undertaking for brownfield companies.”
Andersen said controlling costs on the AWS platform can also be difficult.
“The key is in the details. My concern with AWS is that the pricing is very modular. That can be good if you really know what you’re doing, but it can also be bad if you just use everything and don’t monitor it,” he said.
Andersen said companies are unlikely to move away from GitHub repositories, but may look to competitors for other features.
“It slams the door open,” he said. “They might source agents differently.”
AWS can throw its cloud weight into the agent battle
Spokespeople for GitHub and AWS would not confirm or deny reports published this week that GitHub would run its services on AWS to resolve reliability issues and meet infrastructure capacity demands for agents.
But AWS’ critical mass as the largest cloud hyperscaler by global data center footprint and infrastructure market share could also impact organizations that have not yet developed an AI agent orchestration strategy, analysts said.
“Everyone has problems building new data centers, but if you already have hundreds of data centers, [you can more easily] adding capacity,” said Larry Carvalho, principal consultant at RobustCloud. “AWS may have a secret weapon” [with] The trainium silicon, [and] Because you control the entire stack, you can potentially accelerate enterprise adoption with a lower-cost model. ”
IDC analyst Matthew Flug said AWS Context and Amazon Quick could have an additional advantage given the widespread enterprise use of S3 for storage.
“Many organizations are already connected to AWS in some way and are now expanding to cover that layer while complying with the openness required by the industry with cross-cloud data integration in AWS Context,” said Flug. “In that respect, I think it makes sense for many organizations to continue using AWS for things like Quick.”
Beth Pariseau, senior news writer at Informa TechTarget, is an award-winning IT journalism veteran. Any tips? send an email to her or connect linkedin.
