Should This AI Project Start? A Decision Framework
Compare AI and non-AI options through purpose, people, data, evidence, guardrails, staged tests, total cost, accountability, and a documented proceed-or-stop decision.
An AI idea may need to stop, get more study, face a small test, use a non-AI fix, buy a tool, get built, or grow. There is no sound proof that most AI projects should never start. The goal is a written choice that weighs purpose, other options, proof, effects, duties, and the cost of being wrong.
The NIST AI Risk Management Framework groups optional work under Govern, Map, Measure, and Manage. Its Playbook is not one list for all teams. The Manage guide asks whether a system meets its goal and whether work or use should go on.
Frame the Decision Before Naming AI
Describe the task, people who may feel the effects, current results, problem or chance, owner, limits, and choice the project must aid.
Set the allowed goal and banned uses before you pick a model, vendor, data source, screen, or system design.
Tell apart facts, guesses, forecasts, wishes, and promises. Log who gave each one and when to check it again.
Compare AI With Real Alternatives
Compare no action, process fixes, training, rule changes, search, data study, task automation, bought software, human work, a small AI part, and a larger system. Use the same needs, time span, full-cost base, risks, and pass proof. An AI label is not a reason to say yes or no.
Map People, Data, and Context
Name users, data subjects, non-users, reviewers, staff, customers, workers, and groups that may gain or face harm. Map data source, right to use, quality, fit, private traits, access, vendors, keep time, deletion, safety, ownership, and legal duties. Do not use data just because you have it.
Define Evidence and Guardrails
Set the main result and quality measures, current base, allowed group, test data, error cost, doubt, review steps, and pass line.
Add safety, fairness, privacy, security, access, clear reasons, human review, complaint, staff load, money, and green limits as needed.
Name who can pause, overrule, fix, roll back, or retire the system. Give affected people a way to ask for review or help.
Use Staged Decisions
Steps may include study, a data check, a mock-up, a fair test, a small pilot, watched release, wider use, and retirement. For each step, set entry proof, spend rights, stop rules, owner, review date, and rollback. A demo or test score does not prove a system is safe for live work.
Verify Vendors and Custom Builds
Check current model and product notes, contract, data use, other vendors, locations, access, logs, keep time, deletion, issue duties, test proof, change notice, help, export, and exit. For a custom build, set safe coding, parts, tests, checks, staff, upkeep, and handoff. Buying or building does not remove duty.
Maintain an AI Decision Record
Keep the purpose, other options, proof, group input, dissent, conflicts, approvals, test results, issues, changes, real results, and reasons to go on or stop. Review the choice when the model, data, setting, law, vendor, users, results, or risk changes.
What TTGC Can Support
TTGC can help frame a use case, compare options, map needs and risks, make a mock-up, test, build agreed parts, and set records and checks. TTGC cannot promise approval, truth, savings, use, legal status, revenue, or other AI project results.
Should this AI project move ahead?
TTGC can help frame the choice, compare options, set proof and limits, and plan staged tests. AI and business results are not promised.
Sources
- NIST — AI Risk Management Framework 1.0. https://www.nist.gov/itl/ai-risk-management-framework
- NIST — AI Risk Management Framework Playbook. https://airc.nist.gov/airmf-resources/playbook/
- NIST AI RMF Playbook — Manage. https://airc.nist.gov/airmf-resources/playbook/manage/
- Federal Trade Commission — Privacy and Security guidance. https://www.ftc.gov/business-guidance/privacy-security
