Turn this comparison into procurement-ready diligence
Free shows today's risk. A Stack Audit gives you a citable, verbatim-sourced PDF across your whole AI stack — and flags the moment a vendor's terms change.
SIDE-BY-SIDE · 2 OF 3
Two platforms.
Same questions.
GitHub Copilot
DEVELOPER / CODING
PROMPT
GitHub does not claim ownership over code inputs. Code submitted to Copilot remains the user's property.
OUTPUT
GitHub does not assert ownership over code suggestions. Users are responsible for determining the IP status of generated code.
TRAINING
GitHub Privacy Statement confirms: 'Product Development and Improvement: We use Personal Data to develop and improve our products, services, and technologies, including artificial intelligence and machine learning technologies. This includes improving features, developing new offerings, enhancing safety and security capabilities, and training models. We apply appropriate technical safeguards, including aggregation and de-identification techniques where feasible.' Affiliates including Microsoft may also receive data for 'product development and improvement (including training and improving artificial intelligence and machine learning technologies).' Individual tier: code snippets and telemetry may be used for training (opt-out in IDE settings). Business/Enterprise: training on user code is off by default, no data retention beyond session.
COMMERCIAL
Commercial use is the primary intended use case, particularly for Copilot for Business and Enterprise plans.
208 verified findings across the displayed surfaces.
Terrakotta
WORKFLOW & AUTOMATION
PROMPT
Not explicitly addressed as a distinct surface. Customer Data is licensed to Company under a broad worldwide license (see training_use citation), but the terms do not contain a specific clause addressing who owns prompt inputs. Company exclusively owns the Service and all related IP. Customer Data protection is addressed under confidentiality obligations.
OUTPUT
Not explicitly addressed for AI-generated outputs. However, feedback or suggestions provided by Customer or Authorized Users are subject to an unlimited, irrevocable, perpetual, sublicensable, royalty-free license granted to Company for any purpose, with no compensation obligation. The agreement is non-assignable by Customer without Company's prior written consent, which may limit Customer's ability to transfer output-related rights.
TRAINING
Customer grants Company a worldwide, non-exclusive license to access, use, copy, distribute, perform, and display Customer Data explicitly for 'developing, training, and improving Company's machine learning models, including large language models, and artificial intelligence algorithms.' This license also covers maintaining and updating the Service and sharing data with third-party service providers (e.g., hosting providers) acting on Company's behalf. No opt-out mechanism for training use is specified in the cited terms.
COMMERCIAL
Customer is prohibited from: reverse engineering or decompiling the Service; creating derivative works; using the Service for timesharing, service bureau, or third-party benefit purposes; sublicensing, reselling, or redistributing the Service; and accessing the Service to build a competitive product or service. The Service is classified as a 'commercial item' under FAR section 2.101 and DFAR section 252. Company retains all rights, title, and interest in the Service, including modifications and derivative works.
Not yet scored — verified corpus incomplete.
Empty Comparison Slot
AIRIN SUMMARY
GitHub Copilot carries the most permissive license terms of the set; Terrakotta has no scored posture yet (unrated).FOR ENTERPRISE: GITHUB COPILOT NOT RECOMMENDED · TERRAKOTTA NOT YET SCORED
↳ Compliance surfaces · side by side
⚡ INTERACTIVE RISK COMPARISON HEATMAP
| RISK SURFACE | GITHUB COPILOT | TERRAKOTTA |
|---|---|---|
| PPrompt Ownership | ||
| OOutput Ownership | ||
| TModel Training | ||
| CCommercial Use | ||
| DData Privacy | ||
| $Tier Differences |
* Click any cell in the heatmap matrix to scroll directly to its verbatim policy citations below.
📂 Intellectual Property & Model Training
P
SURFACE 01
Prompt Ownership
GitHub Copilot
MEDHIGH
GitHub does not claim ownership over code inputs. Code submitted to Copilot remains the user's property.
“ “Input” means any content or context provided to an AI Feature, including prompts, attachments, code in your workspace, and conversation history.”
Terrakotta
UNRATED
Not explicitly addressed as a distinct surface. Customer Data is licensed to Company under a broad worldwide license (see training_use citation), but the terms do not contain a specific clause addressing who owns prompt inputs. Company exclusively owns the Service and all related IP. Customer Data protection is addressed under confidentiality obligations.
No citation available
O
SURFACE 02
Output Ownership
GitHub Copilot
MEDHIGH
GitHub does not assert ownership over code suggestions. Users are responsible for determining the IP status of generated code.
“ You grant us and other Users the licenses in Sections D.4–D.8. These licenses apply to Your Content. If Your Content already includes a license granting us the permissions we need, no additional license is required. You will not receive payment for these rights. These licenses e…”
Terrakotta
MED⚠ CAUTIONUNRATED
Not explicitly addressed for AI-generated outputs. However, feedback or suggestions provided by Customer or Authorized Users are subject to an unlimited, irrevocable, perpetual, sublicensable, royalty-free license granted to Company for any purpose, with no compensation obligation. The agreement is non-assignable by Customer without Company's prior written consent, which may limit Customer's ability to transfer output-related rights.
“No rights or licenses are granted except as expressly set forth herein.”
T
SURFACE 03
Model Training
GitHub Copilot
MED⚠ CAUTIONHIGH
GitHub Privacy Statement confirms: 'Product Development and Improvement: We use Personal Data to develop and improve our products, services, and technologies, including artificial intelligence and machine learning technologies. This includes improving features, developing new offerings, enhancing safety and security capabilities, and training models. We apply appropriate technical safeguards, including aggregation and de-identification techniques where feasible.' Affiliates including Microsoft may also receive data for 'product development and improvement (including training and improving artificial intelligence and machine learning technologies).' Individual tier: code snippets and telemetry may be used for training (opt-out in IDE settings). Business/Enterprise: training on user code is off by default, no data retention beyond session.
“ The opt-out right described in this section applies solely to the use of your Inputs and Outputs as described in this section, and does not apply to the license granted in Sections D.4-D.8 or to any Content that does not constitute Input or Output (including Content stored in pu…”
Terrakotta
UNRATED
Customer grants Company a worldwide, non-exclusive license to access, use, copy, distribute, perform, and display Customer Data explicitly for 'developing, training, and improving Company's machine learning models, including large language models, and artificial intelligence algorithms.' This license also covers maintaining and updating the Service and sharing data with third-party service providers (e.g., hosting providers) acting on Company's behalf. No opt-out mechanism for training use is specified in the cited terms.
No citation available
C
SURFACE 04
Commercial Use
GitHub Copilot
MED⚠ CAUTIONHIGH
Commercial use is the primary intended use case, particularly for Copilot for Business and Enterprise plans.
“ By using automated means to access, collect, or otherwise use (“Access”) any publicly accessible Content from the Service for the purpose of developing or training any commercially available artificial intelligence model, machine learning system, or similar technology (a "Commer…”
Terrakotta
MED⚠ CAUTIONUNRATED
Customer is prohibited from: reverse engineering or decompiling the Service; creating derivative works; using the Service for timesharing, service bureau, or third-party benefit purposes; sublicensing, reselling, or redistributing the Service; and accessing the Service to build a competitive product or service. The Service is classified as a 'commercial item' under FAR section 2.101 and DFAR section 252. Company retains all rights, title, and interest in the Service, including modifications and derivative works.
“ 2.1. Usage Restrictions. Customer will not, directly or indirectly: (a) reverse engineer, decompile, disassemble or otherwise attempt to discover the source code, object code or underlying structure, ideas, know-how or algorithms relevant to the Services or any software, documen…”
📂 Data Privacy & Governance
D
SURFACE 05
Data Privacy
GitHub Copilot
MED⚠ CAUTIONHIGH
Individual: training on by default, opt-out available in IDE settings. Business: training off by default, no content retention. Enterprise: same as Business with additional enterprise governance and IP indemnification.
“ We aim to promptly respond to requests in compliance with legal requirements. Please note that we may retain certain data as necessary for legal obligations or for establishing, exercising, or defending legal claims.”
Terrakotta
MED⚠ CAUTIONUNRATED
—
“ In Short: We do not knowingly collect data from or market to children under 18 years of age . ”
$
SURFACE 06
Tier Differences
GitHub Copilot
MEDHIGH
Individual: training on by default, opt-out available in IDE settings. Business: training off by default, no content retention. Enterprise: same as Business with additional enterprise governance and IP indemnification.
“ "Beta Previews" mean software, services, or features identified as alpha, beta, preview, early access, or evaluation, or words or phrases with similar meanings.”
Terrakotta
MEDUNRATED
—
“ 5.3. If Customer believes that Company has billed Customer incorrectly, Customer must contact Company no later than 30 days after the closing date on the first billing statement in which the error or problem appeared, in order to receive an adjustment or credit. Inquiries should…”