GitHub Copilot

GitHub Copilot suggests code, explains repositories, and reviews changes.

Oximy ResearchUpdated 19 September 2026

Overview

A+
Transparency
8 / 10
Trust
95 / 100
Developer ToolsGitHub

Trains on your data

Free
Yes
Paid
No
Opt-out
Yes

Incidents

2
  • mediumInsufficiently protected credentials in GitHub Copilot
  • mediumCritical GitHub Copilot Chat Vulnerability (CamoLeak) Leaked Private Source Code and Secrets
SOC 2ISO 27001GDPR

Assessed SEP 21, 2026 · Checked by hand SEP 21, 2026

GitHub confirms individual-plan training with opt-out, no business or enterprise training, provider ZDR and regional hosting, with two supported patched vulnerabilities.

Trust breakdown

A+

95 / 100

Overall trust score

Data safetyComplianceIncidentsLegal
Data safety100
Compliance92
Incidents85
Legal100

Data and privacy

Trains on your data

Free
Yes
Paid
No
Opt-out
Yes

On by default. Admins can turn it off, which is why it costs less than training with no way out.

Storage regions

United StatesEuropean Union

Sub-processors

Amazon Web Services Inc (AWS)Anthropic PBCCerebras Systems Inc.CloudflareDataDome Solutions, Inc.Elasticsearch, Inc.FastlyFireworks AIFullStory, Inc.Google Cloud Platform (GCP)Hewlett-Packard LimitedLambdaTestMicrosoft (Azure)MicrosoftMoveworksNexMo Inc (aka Vonage)Obsidian SecurityOpenAIOracle America, Inc.PusherTines Automation Inc.Twilio (SendGrid)xAIZendesk

Data retention

GitHub maintains a zero data retention agreement with OpenAI.

Deletion: Following the completion of the Services, to the extent that GitHub is a Processor and unless prohibited by law, GitHub will delete or return all the Customer Personal Data to You, whichever you elect, and delete existing copies in accordance with our retention and deletion policy.

Privacy policyTermsAcceptable useTrust center

Incidents

Powered by

GPT-4.1Claude Sonnet 3.5Claude Sonnet 3.7Claude Sonnet 3.7 ThinkingClaude Sonnet 4Claude Opus 4Claude Opus 4.1GPT-5GPT-5 minio3o3-minio4-miniGemini 2.5 ProGPT-5 miniGPT-5.3-CodexGPT-5.4GPT-5.4 miniGPT-5.4 nanoGPT-5.5GPT-5.6 LunaGPT-5.6 SolGPT-5.6 TerraClaude Fable 5Claude Fable 5.1Claude Haiku 4.5Claude Opus 4.5Claude Opus 4.6Claude Opus 4.7Claude Opus 4.8Claude Opus 4.8 (fast mode)Claude Opus 5Claude Sonnet 4.5Claude Sonnet 4.6Claude Sonnet 5Gemini 3.1 ProGemini 3.5 FlashGemini 3.6 FlashGemini 3.7 FlashGemini 3.8 FlashMAI-Code-1-FlashMAI-Code-1.1-FlashRaptor miniKimi K2.7 CodeKimi K3Grok 4.5Grok 4.6
Proprietary model

GitHub Copilot uses a multi-model architecture, defaulting to GPT-4.1 and offering a wide range of advanced models from OpenAI, Anthropic, Google, Microsoft, Moonshot AI, and xAI. It also utilizes its own fine-tuned models and a system called Project HydraFusion for orchestrating different models.

Integrations and access

Integrations

API: YesOpenAIMicrosoft Foundry / Azure OpenAIAnthropicOllama

Available on

Web
iOS
Android
macOS
Windows
Linux
Extension
CLI

Pricing

Free

$0 USD

2,000 completions per month, access to Haiku 4.5, GPT-5 mini, and more, Copilot CLI, Community Support

Pro

$10 USD per user / month

Everything in Free, plus access to Cloud agent and code review, unlimited code completion and next edit suggestions, access to 3rd party agents (Claude Code and Codex), model selection, $15 monthly total credits

Pro+

$39 USD per user / month

Everything in Pro, plus access to premium models including Opus, audit logs, 4x+ included usage than Pro, $70 monthly total credits

Max

$100 USD per user / month

Everything in Pro+, plus priority access to new models and features, 2.9x+ included usage than Pro+, $200 monthly total credits

Copilot Business / Copilot Enterprise

Custom

Copilot Business: $19 USD per granted seat per month, 1,900 GitHub AI Credits per user per month. Copilot Enterprise: $39 USD per granted seat per month, 3,900 GitHub AI Credits per user per month. Both offer access to premium models, with Enterprise having priority access. Usage is measured in AI credits, with additional usage charged at $0.01 USD per AI credit beyond the included pool. Admins can set budgets.

View pricing page

Prices as listed on SEP 21, 2026; check the vendor's page.

Moat and openness

7/ 10
Moat 7 out of 10

Moat

  • Proprietary Model
  • Proprietary Data
  • Network Effects
  • Switching Costmoderate
  • Unique UX
  • Distributionstrong

GitHub Copilot exhibits a strong moat, primarily driven by its deep integration within the developer workflow, extensive feature set, and significant market share. While it leverages a multi-model architecture, its proprietary orchestration (HydraFusion) and fine-tuning, coupled with access to GitHub's vast code data, create a unique and sticky user experience. The high switching costs for organizations and the strong distribution through GitHub's platform further solidify its defensibility.

6/ 10
Openness 6 out of 10

Openness

  • Open model weights
  • Published research
  • Open source contributions
  • Transparency reports
  • Public safety evals

GitHub demonstrates a moderate level of openness and transparency. They publish comprehensive transparency reports through their dedicated Transparency Center, detailing requests for user information, government and DMCA takedowns, and automated detection. They also provide structured data files for these reports in a public GitHub repository, encouraging further analysis. While they are transparent about their content moderation and legal request processes, information regarding the openness of their AI model weights and public safety evaluations is not readily available.

Timeline

  1. 2023-03GitHub Copilot X is revealed, introducing chat interfaces, voice commands, and pull request summaries.
  2. 2022-12GitHub Copilot for Business is launched, providing enterprise-grade security and license management.
  3. 2022-06GitHub Copilot becomes generally available to individual developers as a subscription service.
  4. 2021-06GitHub and OpenAI announce the technical preview of GitHub Copilot for Visual Studio Code.

Company

GitHub

Founded
2008
HQ
San Francisco, California, U.S.

Popularity

1.5M

monthly visits

high tierlow risk

Notable customers

AMDAstraZenecaTrimbleASOS

GitHub Copilot is the largest AI coding assistant by paid user base and developer awareness, holding approximately 42% market share among paid AI coding tools.

Value and ROI

What it should move: It shortens implementation, reviews, and repetitive coding work.

Best fit

It fits teams that want assistance inside GitHub's source control and reviews.

The catch

Generated code can repeat insecure or unsuitable patterns. Teams still need repository rules, licence checks, tests, and pull-request approval.

  • ASOS: Uses GitHub Copilot to navigate codebases, prepare features and conduct initial code reviews.[1]
  • Duolingo: Uses GitHub Copilot for code consistency, boilerplate and routine engineering work.[2]

In our stacks for

Engineering

References

  1. 01ASOS: how it uses GitHub CopilotGitHubVendor-published
  2. 02Duolingo: how it uses GitHub CopilotGitHubVendor-published
  3. 03Privacy policyGitHubVendor-published
  4. 04TermsGitHubVendor-published
  5. 05Acceptable useGitHubVendor-published
  6. 06Trust centerGitHubVendor-published
  7. 07Pricing pageGitHubVendor-published
  8. 08Company siteGitHubVendor-published
  9. 09SOC 2GitHubVendor-published
  10. 10GDPRGitHubVendor-published
  11. 11Insufficiently protected credentials in GitHub CopilotMicrosoft Security Response CenterIndependent
  12. 12Critical GitHub Copilot Chat Vulnerability (CamoLeak) Leaked Private Source Code and SecretsLegit SecurityIndependent