TargetZero
AI that answers to your agency.
Zero is TargetZero’s AI assistant. Your team asks it for what it needs, and it answers from your agency’s knowledge. Here’s how we keep it accountable.
A human decides
Zero prepares the work. A person on your team decides what to do with it. TargetZero only reads from your systems. It doesn’t change records in them. Automations, when they come, will be optional, and your agency will be able to turn them off.
What Zero won’t do
Zero doesn’t predict who will commit a crime, and it doesn’t score or rank people.
Built on your agency’s knowledge
Zero works from your records, linked across your systems. It counts the way your agency counts: your list of major crimes, your zones, your policies. General-purpose AI doesn’t know any of that.
Every answer shows its sources
Each answer links to the records behind it. Open them and check the work. Where records disagree, TargetZero highlights the conflict for a person to review instead of guessing.
The same rules as your people
Zero follows the same access rules as the person asking. No one sees a record through Zero that they couldn’t open themselves. Automations will follow the same rules as your team.
Everything is logged
Every action by people and by Zero is logged, with who did it and when. Your agency can request that log at any time.
Your data stays yours
It’s never sold, never shared and never used to train AI models. If you leave, you get an export of your data, and we delete it from TargetZero within 30 days.
Guided by NIST
We follow the NIST AI Risk Management Framework as our guide. TargetZero is hosted in AWS GovCloud (US) and built to the FBI CJIS Security Policy’s requirements.