Comparing AI Assistant, AI as One of the Reviewers, and AI as the Only Reviewer in EasySLR

Comparing AI Assistant, AI as One of the Reviewers, and AI as the Only Reviewer in EasySLR

EasySLR utilises advanced artificial intelligence to streamline and enhance the article screening process. Within the platform, AI can be configured in three different ways depending on your review methodology and project requirements:
  • AI Assistant
  • AI as One of the Reviewers
  • AI as the Only Reviewer
While all three modes leverage the same underlying AI capabilities, they serve different purposes within the review workflow.

Choosing the Right AI Configuration
EasySLR provides complete flexibility in determining how AI is used at each screening stage.
Users can choose whether AI should:
  • Act as an Assistant supporting human reviewers
  • Function as one of the Reviewers alongside human reviewers
  • Serve as the Only Reviewer for a particular stage or all the stages
These configurations can be applied independently at the:
  • Title & Abstract Screening stage
  • Full-Text Screening stage
This allows teams to tailor AI involvement according to project complexity, resource availability, and methodological requirements.

AI Assistant
What It Does
AI Assistant focuses on helping reviewers make faster and more informed decisions.
It:
  • Reviews articles using predefined screening criteria
  • Generates inclusion/exclusion suggestions
  • Provides AI-generated rationale and supporting notes
  • Prioritizes likely inclusions
  • Reduces manual effort during screening
Key Benefit
The reviewer remains fully in control while benefiting from AI-generated recommendations.

AI as One of the Reviewers
What It Does
AI Reviewer acts as an independent reviewer alongside human reviewers.
It:
  • Reviews articles independently
  • Applies screening criteria consistently
  • Generates inclusion/exclusion decisions
  • Participates in conflict generation
  • Supports quality control processes
Key Benefit
Combines AI consistency with human expertise, improving reliability and reducing reviewer variability.

AI as the Only Reviewer
What It Does
In this configuration, AI performs screening independently without requiring human reviewer decisions for the selected stage.
It:
  • Reviews all assigned records automatically
  • Applies protocol criteria consistently
  • Generates final screening decisions
  • Progresses articles through the workflow
Key Benefit
Provides maximum efficiency and scalability, making it ideal for projects requiring rapid processing of large evidence sets.

Feature Comparison:
Aspect
AI Assistant
AI as One of the Reviewers
AI as the Only Reviewer
Purpose and Function
Optimises the screening workflow by providing suggestions and rationales to assist reviewers.
Complements human reviewers with an independent assessment and quality control layer.
Performs screening independently without requiring human reviewer decisions at the selected stage.
Role in Workflow
Supports reviewer decision-making. Final decisions remain with the reviewer.
Functions as one of the reviewers in a dual-review workflow.
Functions as the sole reviewer for the selected stage.
Screening Decisions
Provides recommendations only. Does not make final decisions.
Makes independent screening decisions alongside human reviewers.
Makes and records final screening decisions automatically based on Project Inclusion/Exclusion Criteria.
Features and Workflow
Generates inclusion/exclusion suggestions and supporting rationale using predefined criteria such as PICOS and hierarchy. Prioritises inclusions to improve efficiency.
Conducts independent evaluations using predefined criteria such as PICOS and hierarchy. Participates in conflict generation and resolution workflows.
Independently evaluates all records using predefined criteria such as PICOS and hierarchy and progresses articles through the workflow without manual intervention.
Quality Control
Reviewer validates AI suggestions before making decisions.
Provides an additional review layer that improves consistency and reliability.
Suitable for projects where fully automated screening is desired and human review is not required at that stage.
Setup and Configuration
Enable AI Notes and Suggestions within the AI Copilot section.
Enable AI as a Reviewer within the AI Copilot section. If configured for an SLR workflow, AI suggestions and notes are hidden to avoid reviewer bias.
Configure AI as the only reviewer by setting the required number of reviewers accordingly and enabling AI Reviewer functionality without assigning human reviewers to that stage.
Human Involvement
High. Human reviewers make all final decisions.
Moderate. Human reviewers work alongside AI.
Minimal. AI performs all screening decisions for the configured stage.
Benefits
Saves time, reduces manual effort, highlights relevant articles, and improves workflow efficiency.
Enhances accuracy, consistency, and quality control through independent assessments.
Maximises screening efficiency, significantly reduces manual workload, and accelerates project completion.
Best Use Cases
Projects where reviewers want AI support while maintaining complete control over decisions.
Systematic reviews requiring independent dual-review workflows.
Rapid reviews, large-scale evidence surveillance, living reviews, or projects where automated screening is preferred.

Summary
AI Assistant: Focuses on supporting human reviewers by providing recommendations, rationale, and prioritisation while keeping decision-making entirely in human hands.
AI as One of the Reviewers: Acts as an independent reviewer alongside human reviewers, improving consistency, quality control, and methodological rigor.
AI as the Only Reviewer: Performs screening autonomously, enabling fully automated screening workflows for teams seeking maximum efficiency and scalability.

Note
Users can configure AI differently at each screening stage. For example:
  • AI Assistant at Title & Abstract Screening
  • AI Reviewer at Full-Text Screening
or
  • AI as the Only Reviewer at Title & Abstract Screening
  • Human Reviewers at Full-Text Screening
This flexibility allows teams to adopt the level of AI involvement that best aligns with their review methodology and project requirements.