AI Protest Risk Assessment Playbook
Your firm submitted a proposal on a $12M HHS contract and was not selected. The award went to a competitor at a price 18% lower than yours. Your BD team believes the awardee cannot deliver at that price and that the technical evaluation may have been flawed. You are considering a GAO protest.
When to use this playbook
- Use this playbook when the decision looks like the situation above: Your firm submitted a proposal on a $12M HHS contract and was not selected.
- It is a fit when you have source files in hand and need a structured, reviewable analysis — not a generic chat answer about "Protest Risk Assessment".
- Do not use it as a substitute for licensed, legal, clinical, or authorized official judgment in the domain.
What you'll need
- Your firm's proposal (redacted for proprietary data)
- The debriefing notes from HHS (written and oral)
- The solicitation evaluation criteria
- Any award information available (SAM.gov, FPDS)
- Your firm's technical score and the stated reasons for non-selection
Attachments: Documents (Documents)
The Prompt
You are a federal procurement attorney assessing the viability of a GAO protest for a $12M HHS contract. I am attaching: Work only from the attached source files. If a conclusion is not supported, say so. Produce: 1. Assess the protest grounds: was the evaluation unreasonable, was there a solicitation ambiguity, was the best value tradeoff documented, or was there an OCI issue? 2. Identify the specific regulatory or FAR provision violated if the non-selection was improper—a protest needs a legal basis, not just a feeling the decision was wrong. 3. Calculate the probability of success based on GAO sustain rates for the type of protest ground identified. 4. Assess the business risk: what is the cost of a protest, the timeline, and the likelihood of corrective action vs. full award if sustained? 5. Tell me whether to protest, request additional debriefing, or move on—and if we protest, what documents to request under the protective order. Call out where independent models are likely to disagree, and list follow-up documents a reviewer should request.
What to expect
- Protest grounds analysis with FAR citation
- GAO sustain rate estimate for identified grounds
- Business risk analysis (cost, timeline, outcomes)
- Protest/no-protest recommendation
- Document request list for protective order
Review before you act
- Validate this output against source files before relying on it: Assess the protest grounds: was the evaluation unreasonable, was there a solicitation ambiguity, was the best value tradeoff documented, or was there an OCI issue?.
- Validate this output against source files before relying on it: Identify the specific regulatory or FAR provision violated if the non-selection was improper—a protest needs a legal basis, not just a feeling the decision was wrong.
- Validate this output against source files before relying on it: Calculate the probability of success based on GAO sustain rates for the type of protest ground identified.
- Validate this output against source files before relying on it: Assess the business risk: what is the cost of a protest, the timeline, and the likelihood of corrective action vs. full award if sustained?.
- Confirm every cited figure, date, counterparty, or requirement against the attached originals — models compress and can drop a qualifier.
- Treat disagreement between models as a review item, especially on classification, materiality, and recommended next action.
- Do not authorize an operational, clinical, legal, credit, or enforcement action solely because the models agree.
Why compare models on this
For Protest Risk Assessment, running the same attachments across independent models is useful because the hard part is classification and completeness, not fluency. The workflow is already designed to surface protest grounds analysis with far citation; gao sustain rate estimate for identified grounds; business risk analysis (cost, timeline, outcomes); protest/no-protest recommendation. Those are comparison artifacts — they only exist if more than one model runs. Models split on whether a requirement is mandatory, how to score a differentiator, and protest likelihood. Those splits should be resolved before color-team review, not after submission.
See governed multi-model AI on your own prompt
Compare GPT-5, Claude, and Gemini side by side, with human review and a decision record built in.

