Live
Local AWS Emulator reshapes the verification loop for coding agentsClaude Enterprise Expansion: Architecture and Risk Implications for AI‑Powered SaaSDeploying OpenFold3 with SageMaker Asynchronous Inference: Architecture and Ops GuideClaude Fable 5.2 reduces safeguard blocks, speeds up multi‑turn coding, and lowers token pricingLLM coding cost: Claude Fable 5.2’s lower list price hides higher token spend than GPT‑6 AstraSparse Autoencoders Enable Direct Tool Choice Auditing in LLM Agents via Kiji InspectorIntegrating Google’s Gemini Decision API: Faster, Lower‑Cost Bounded Choices for AI WorkloadsCreating a Continuous Model‑Improvement Loop Using SageMaker HyperPod Data CaptureLocal AWS Emulator reshapes the verification loop for coding agentsClaude Enterprise Expansion: Architecture and Risk Implications for AI‑Powered SaaSDeploying OpenFold3 with SageMaker Asynchronous Inference: Architecture and Ops GuideClaude Fable 5.2 reduces safeguard blocks, speeds up multi‑turn coding, and lowers token pricingLLM coding cost: Claude Fable 5.2’s lower list price hides higher token spend than GPT‑6 AstraSparse Autoencoders Enable Direct Tool Choice Auditing in LLM Agents via Kiji InspectorIntegrating Google’s Gemini Decision API: Faster, Lower‑Cost Bounded Choices for AI WorkloadsCreating a Continuous Model‑Improvement Loop Using SageMaker HyperPod Data Capture
GitHub

Copilot model deprecation forces pipeline updates and policy reviews

AI SummaryPowered by AI

GitHub will retire a set of Copilot models on December 14 2026, replacing them with newer alternatives. The change forces AI‑engineers and platform teams to adjust integrations, model selections, and policy configurations to avoid service disruption.

Effective December 14 2026, GitHub is removing several Copilot models from all experiences—including Chat, inline edits, ask, agent modes, and code completions. Practitioners who embed Copilot in CI pipelines, custom tooling, or enterprise settings must switch to the suggested replacements before that date, or risk losing model access.

Deprecation timeline and replacement models

  • Claude Fable 5 → Claude Fable 5.2
  • Claude Haiku 4.5* → Claude Haiku 5.5 (note: 4.5 remains for individual annual‑plan subscribers)
  • Gemini 3.8 Flash → Gemini 4 Hydrogen
  • Gemini 4 Argon → Gemini 4 Carbon
  • Grok 4.6 → Grok 4.7
  • Kimi K3 (no direct alternative listed)

All deprecations take effect on the same day, 2026‑12‑14.

Impact on CI/CD and automation

Any automation that calls Copilot APIs, uses model identifiers in configuration files, or relies on default model behavior will encounter failures once the old identifiers are removed. Teams should audit scripts, GitHub Actions, and internal services for hard‑coded model names and replace them with the new identifiers. Because the platform does not automatically purge the deprecated models, no explicit removal step is required, but calls to the old names will no longer resolve.

Policy and configuration considerations

Copilot Enterprise and Business customers receive the suggested alternatives automatically through the global default model enablement. However, administrators can disable the global default or explicitly block a model via the Copilot settings model policies. If an admin has turned off the default, they must manually enable the replacement models to make them visible in the model selector. Individual annual‑plan subscribers retain access to Claude Haiku 4.5, so their workflows may continue unchanged if they do not opt into the newer version.

Related CloudNinjas coverage: AI engineering.

What This Means For Practitioners

  • Audit all code‑generation integrations for references to the deprecated model names.
  • Update configuration files, scripts, or action workflows to use the new model identifiers before 2026‑12‑14.
  • Verify that global default model enablement is active for Enterprise/Business tenants, or adjust model policies to grant access to the replacements.
  • For individual annual‑plan users, confirm whether the existing Claude Haiku 4.5 meets current quality needs or plan a migration to Haiku 5.5.
  • Contact your GitHub account team if you need guidance on policy changes or encounter unexpected behavior after the deprecation date.
Originally published atGitHub Changelog