IceShoreRadar AI × Software Maintenance

IceShore Radar AI × Software Maintenance

October 2026

[GOOD NEWS]

5.4%86.6%

COBOL migration accuracy

OpenAI says its newest model now gets COBOL migrations right 86.6% of the time

OpenAI says GPT-6.1 Sol correctly migrated COBOL programs to languages such as Java or Python without changing their behavior 86.6% of the time at a low reasoning setting, up from 5.4% for GPT-5.6 Sol, and 97.9% with more reasoning. The tests are OpenAI's own. BleepingComputer notes that real migrations would still need expert review and testing.

・ US tech news site · report

Source: BleepingComputer (bleepingcomputer.com)

Photo: Markus Spiske markusspiske / Wikimedia Commons, CC0

[UPDATE]

Error budgets

allowed failure rate, now for AI agents

Grafana Labs proposes managing AI agents with service-level objectives and error budgets

InfoQ reports Grafana Labs wants teams to measure AI agent behavior, such as hallucination and task fulfilment, and manage it with service-level objectives and error budgets. Its example target: 95% of conversations judged as fulfilled over 30 days, with the rest as the error budget.

・ US observability software company · report

Source: Grafana Labs (infoq.com)

Photo: Alan Levine from United States / Wikimedia Commons, CC BY 2.0

[UPDATE]

50%+

shorter investigation phase

HERE expects AWS Kiro agents to cut its task investigation phase by more than half

On an AWS blog, map company HERE Technologies said it uses multiple Kiro agents to turn customer requirements into Jira tickets that keep the traceability automotive audits demand. Based on early internal measurements, it expects to cut the investigation phase for new tasks by more than 50%.

・ Dutch mapping company · AWS blog

Source: HERE Technologies (aws.amazon.com)

Photo: bizmac (Flickr) / Wikimedia Commons, CC BY 2.0

[GOOD NEWS]

2 months1 month

planning time per project

LY Corp. says AI design help cut planning on its 27-year-old shopping site to one month

LY Corp. said on its tech blog that AI agents now draft requirements and designs for its Japanese shopping site, a service with over 3,000 repositories and 27 years of history. The team first organized domain knowledge and system diagrams for the AI, and senior engineers review each stage. Planning time fell from about two months to roughly one.

・ Japanese internet giant · official blog

Source: LY Corporation (techblog.lycorp.co.jp)

Photo: Syced / Wikimedia Commons, CC0

[UPDATE]

38

functions refactored and redeployed

CyberAgent used generative AI to review about 200 live Lambda functions and refactor 38

Japanese internet company CyberAgent said on its engineering blog that it used generative AI to catalog about 200 AWS Lambda functions, small cloud programs running in production and managed by multiple people. It refactored and redeployed 38 of them. Risky steps such as production deploys required explicit approval from a human engineer.

・ Japanese internet company · tech blog

Source: CyberAgent (developers.cyberagent.co.jp)

Photo: Syced / Wikimedia Commons, CC0

[UPDATE]

Failure drills ×AI root cause

AI watches failure drills

AWS has its DevOps Agent watch resilience tests and report root causes and impact

Amazon Web Services said Oct. 15 that Resilience Hub, its service for checking how applications withstand failures, now starts a DevOps Agent investigation to watch each resilience test and report symptoms, root cause, impact and recovery actions.

・ Major US cloud provider · announcement

Source: Amazon Web Services (aws.amazon.com)

Photo: Rsparks3 / Wikimedia Commons, CC0

[GOOD NEWS]

near 0%50%

orgs using low-touch AI in IT ops

Gartner expects half of organizations to use low-touch AI agents in IT operations by 2030

Research firm Gartner said on Oct. 26 that by 2030, 50% of organizations will use AI agents that need minimal human involvement in IT operations, up from nearly 0% in 2026, Japan's ITmedia reported. In a Gartner survey, 58% of IT operations leaders said moving AI into production was an important priority.

・ US research firm · forecast via ITmedia

Source: Gartner (itmedia.co.jp)

Photo: Olivier Bonin/SLAC National Accelerator Laboratory / Wikimedia Commons, CC BY 4.0

[UPDATE]

17%40%+

pull requests started via agents

GitHub says AI agents now kick off over 40% of pull requests, up from 17% in March

GitHub's annual Octoverse report says that as of September, more than 40% of pull requests, the proposed code changes submitted for review, began with a developer directing an AI agent, up from 17% in March. The report describes code review as the way developers shape what the agents return.

・ US developer platform · annual report

Source: GitHub (github.blog)

Photo: GitHub Chief Legal Officer Shelley McKinley at Web Summit Rio, April 2025. Vaughn Ridley/Web Summit via Sportsfile / Wikimedia Commons, CC BY 2.0

[UPDATE]

Terraform ×root-cause tracing

failures across team stacks

AWS demonstrates DevOps Agent tracing a failure caused by another team's Terraform change

Amazon Web Services published a walkthrough of AWS DevOps Agent, its AI operations tool, tracing a failure across separately managed Terraform stacks. In the example, an infrastructure team's tighter network rule cuts an app team's function off from its database, and the agent uses alarms, dependencies and change logs to point to the likely cause.

・ US cloud provider · tech blog

Source: Amazon Web Services (aws.amazon.com)

Photo: Ajay Suresh / Wikimedia Commons, CC BY 2.0