AI Transparency Statement

AI functionality within Jump EHR is used to assist with:

• Drafting and structuring clinical notes
• Summarising consultation information
• Supporting document and letter generation
• Formatting free text into structured records
• Converting spoken clinical dictation into text (voice transcription)

AI is used to improve efficiency and consistency in documentation. It does not replace clinical expertise.

AI features in Jump EHR do not:

• Provide diagnoses
• Recommend specific treatments
• Make prescribing decisions
• Generate Clinical Decision Support rules
• Replace deterministic safety alerts
• Operate as an autonomous clinical decision-maker

Clinical Decision Support (CDS) within Jump EHR is rule-based and deterministic. AI systems are separate from these safety checks and do not generate or modify CDS logic.

All AI-generated content requires human review.

• Clinicians remain responsible for all clinical decisions
• AI outputs must be reviewed, edited and approved before use
• AI suggestions do not automatically update the clinical record
• Structured extractions and drafts require explicit user confirmation

AI acts as an assistive drafting tool, not an authority.

AI features operate within the context of the customer's environment.

• AI processes data only to provide the requested functionality
• Customer data is not used to train public or shared AI models
• AI functionality can be disabled at organisational level
• Usage controls and limits are in place to manage system behaviour

AI processing is subject to the same data protection controls as other platform features.

Jump EHR uses large language model services provided by external AI providers. Specific models may change over time as technology evolves. Models are configured for assistive drafting and structuring tasks, not clinical decision-making.

AI outputs are based on patterns in the input provided during use. AI does not have independent awareness of individual patients beyond the information entered.

AI responses may contain inaccuracies, omissions or incorrect inferences. They should be treated as draft assistance, not factual conclusions.

AI systems can:

• Misinterpret ambiguous inputs
• Produce incomplete summaries
• Generate plausible but incorrect statements
• Miss context not explicitly recorded

Because of this, clinical review is always required.

Clinical Decision Support in Jump EHR operates using deterministic, rule-based logic derived from structured clinical data and configuration rules.

AI features may help present or explain information, but they do not create, modify or execute CDS rules. Safety alerts are not AI-generated.

AI use within Jump EHR is subject to ongoing monitoring and improvement, including:

• Review of user feedback
• Identification of error patterns
• Updates to prompts and configurations
• Separation between AI features and safety-critical logic

Healthcare organisations control whether AI features are enabled. AI tools can be disabled where they are not appropriate for a workflow.

AI functionality evolves over time. Improvements are made to increase reliability and clarity while maintaining human oversight and data protection safeguards.

For questions about AI use in Jump EHR:

Email: hello@usejump.co.uk