Effective date: 1 June 2026
1. Nature of AI outputs
Stactiq produces outputs — session structures, set piece configurations, corrective plans — through AI inference governed by Customer-defined Methodology DNA and constrained by a Hallucination Firewall that rejects generic, unanchored coaching content.
Governance reduces but does not eliminate the possibility of outputs that do not reflect a Customer's intended doctrine. All system outputs require human review by a qualified coaching professional before operational deployment.
2. Human responsibility
Stactiq is an infrastructure layer for methodology governance. It is not an autonomous coaching decision system. The professional judgement of a qualified technical director, academy director, or head coach is required at every deployment point. Stactiq does not replace that judgement.
Customers are responsible for:
- Verifying that outputs align with the intended game model before distribution to coaching staff.
- Ensuring that Methodology DNA is current and reflects actual club doctrine.
- Exercising professional oversight over all corrective plans produced by Match Reconstructor before inclusion in training programmes.
3. Restricted use cases
Stactiq outputs must not be used:
- As the sole basis for player selection, contract, or development decisions.
- To produce documentation that misrepresents actual training content to governing bodies.
- In contexts where automated AI output carries legal or regulatory weight without human review.
- To evaluate the performance of third-party AI systems.
4. Model limitations
The Session Engine, Set Piece Studio, and Match Reconstructor operate within the constraints of:
- The precision of the Customer's encoded Methodology DNA. Vague or incomplete DNA produces outputs of reduced specificity.
- The quality of match data submitted to Match Reconstructor. Outputs are only as reliable as the diagnostic inputs.
- The inherent non-determinism of large language model inference. Identical inputs may produce varied outputs.
Stactiq's Hallucination Firewall rejects outputs containing generic coaching language not traceable to encoded doctrine. This substantially reduces but does not eliminate output variance.
5. Customer accountability
Customers accept full operational responsibility for outputs they deploy to coaching staff. Stactiq Inc. is not liable for footballing decisions made on the basis of system output.
6. Safety and monitoring
Stactiq monitors inference operations for patterns inconsistent with expected usage. Anomalous usage — including attempts to produce outputs circumventing Methodology DNA governance — is flagged for review and may result in access restriction.
7. Model routing disclosure
AI inference operations within Stactiq are routed to third-party model providers. A current list is available upon request to security@stactiq.com. Routing decisions are made by Stactiq's infrastructure based on availability and capability requirements. Customers do not control model routing. All providers are contractually prohibited from using Customer Content for model training.