AI Output Disclaimer

What stormbraining does

Stormbraining produces commentary on a submitted idea using a panel of AI personas powered by Anthropic's Claude API. The panel runs in three rounds — Round 1 surfaces positives and improvements, Round 2 surfaces risks and reasons it might fail, Round 3 produces a synthesis with prioritized improvements, prioritized risks, and contextual resource recommendations.

What stormbraining is not

The output of Stormbraining is AI-generated commentary, not professional advice. Specifically, it is not:

• Legal advice. Consult a qualified lawyer for legal questions.
• Financial or investment advice. Consult a licensed advisor.
• Medical, regulatory, or safety guidance.
• A guarantee of success or failure for any business idea.

Limits we are honest about

• Hallucinations happen. AI panels may state plausible-sounding facts, statistics, framework attributions, or market sizes that are wrong. We instruct the experts to avoid citing specific numbers unless certain, but errors slip through.
• No real-time market data.The panel does not have access to live market data, current company financials, recent news, or your specific competitors' positioning unless you provide it in the idea text.
• No personalized due diligence. The panel does not know your team, your runway, your investors, or your prior history. It reasons from the idea text alone.

How to use the output well

Treat each expert's output as a useful starting point — a list of things worth investigating — not as a verdict. The synthesis prioritizes the strongest signals across all eight experts; even there, verify load-bearing claims with primary sources before acting.

Reporting inaccurate output

Each expert card has a "Report inaccurate" link. Use it. We use these reports to tune the prompts.