AI Disclaimer

How proposals are generated

When you fill in the brief form and press “Generate,” ProposalForge sends your prompt to Anthropic's Claude model. Claude returns a structured draft (sections, suggested pricing, terms) which we display in the editor. Subsequent “Regenerate” actions follow the same flow on a per-section basis.

We do not pre-screen, fact-check, or verify the output. The model does not have access to the live web, your previous proposals, or your client's history.

What AI can get wrong

• Prices, hours, and timelines. The suggested pricing table is a structural placeholder, not a quote. Always replace it with your real numbers.
• Names and facts. The model can confuse client names, project scopes, or industry terminology. Cross-check against your brief.
• Legal claims. Generated text may make warranty, guarantee, or compliance claims you cannot honor. Edit or remove them.
• Confidential or proprietary data. The model may output something that resembles confidential material. If you recognize anything that should not be in the document, remove it.
• Currency and tax. Default currency and tax rates are illustrative. Set the right ones for your jurisdiction.
• Hallucinations. The model can fabricate credentials, certifications, awards, or case studies. Never present generated text as proof of anything you have not independently verified.

What ProposalForge is not

• Not legal advice. Generated terms and conditions are templates. Have a lawyer review the final wording for your jurisdiction and engagement type.
• Not tax or financial advice. Tax rates and payment schedules are illustrative.
• Not a professional services contract drafting tool. Complex engagements, regulated industries, or deals over a significant value should have an attorney-drafted master services agreement underneath.

Your obligations

• Review every section before sending.
• Replace AI-suggested pricing with your real pricing.
• Verify any factual claims about your business or services.
• Do not include another party's confidential information in a brief unless you have permission to do so.
• Comply with disclosure laws — some jurisdictions require you to inform recipients that AI was used.

How we use your data

Your prompts and generated content are processed by Anthropic under their commercial terms. Anthropic does not retain or train on the data. ProposalForge does not use your content to train any model. See our Privacy Policy for the full data flow.

Limitation of liability

The Terms of Service govern liability for AI-generated output. To the maximum extent permitted by law, ProposalForge is not liable for losses arising from inaccurate AI output, including lost deals, contractual disputes with your clients, or damages caused by your reliance on a generated draft.

Contact

Questions about AI usage: [email protected].