Policy and protocols
Policy text and testing protocols your teams can follow.
Syntony / Services / Governance advisory
Service
Governance advisory Turn test results into decisions people own
Findings change nothing until someone with authority acts on them. We help organizations write AI safety policies and testing protocols, set thresholds and decision rights, and connect red team and evaluation results to the controls and regulations they affect.
- Track record
- Advised three major multinational technology organizations on safety policies, and a leading global bank on cyber risk from advanced AI models.
- Who runs it
- Core team, with policy, legal and technical specialists working side by side.
- Format
- Policy drafting, reviews, workshops and ongoing advisory.
First-person walkthrough
Executive floor
Walk the site in first person and run a red team test at each station: inspect, disrupt, limit, retest. 4 stations, about five minutes.
- Board
- Risk office
- Engineering
- Compliance
Illustrative environment
01 / What we test
What we advise on
Safety policies
Frontier safety frameworks and AI policies, written or revised so each commitment can be tested.
Testing protocols
Model-level and feature-level protocols that say what gets tested, when and by whom.
Thresholds and decision rights
Clear thresholds, and named people with authority to pause, approve or escalate.
Regulatory mapping
How findings bear on the EU AI Act, the NIST AI RMF, ISO/IEC 42001 and sector rules.
Vendor and acceptance reviews
What to ask for, and check, before you buy or deploy a third-party model or agent.
Board and executive briefings
Short, plain readouts on AI risk for the people who carry it.
See an example campaign plan
A fictional plan for a company preparing its first AI safety policy. Each engagement sets its own rows, phases and depth.
Example campaign planFictional plan for illustration
| Discovery | Draft | Pressure test | Adoption | First review | |
|---|---|---|---|---|---|
| Policy | |||||
| Commitments and scope | Primary focus | Primary focus | Covered | Light check | Covered |
| Thresholds | Light check | Primary focus | Primary focus | Light check | Covered |
| People | |||||
| Decision rights | Covered | Covered | Primary focus | Primary focus | Light check |
| Board briefing | Not applicable | Light check | Light check | Primary focus | Covered |
| Evidence | |||||
| Mapping to regulation | Light check | Covered | Light check | Covered | Primary focus |
| Links to test results | Light check | Covered | Primary focus | Light check | Primary focus |
- Primary focus
- Covered
- Light check
- Not applicable
02 / What you keep
What stays with your team
Decision rights map
Who can pause, approve or escalate, and when.
Regulatory crosswalk
Findings and controls mapped to the rules that apply.
Review calendar
When each part gets revisited, and what triggers an early review.
Where we draw the line
We aren't your lawyers. Where regulation is involved we work alongside your legal team, and we say where their judgment is needed.
Discuss your system
Tell us about the system and the decision ahead. We aim to reply within two business days.