Gallup is committed to continuously enhancing Gallup Access with the latest technologies, including artificial intelligence (AI) and its disciplines, to deliver smarter tools, faster insights and an improved user experience.
AI is a broad field encompassing systems that simulate human intelligence, such as understanding language or recognizing patterns. Machine learning (ML) is a field of AI that enables systems to learn from data and improve over time without explicit programming. Generative AI (Gen AI), a newer field of AI, uses advanced models to create new content, such as text, conversations and summaries, based on learned patterns.
Gallup Access incorporates these technologies in the following ways to create intelligent, scientifically grounded and ethically developed tools that deliver meaningful outcomes:
Gallup AI (Gen AI)
Gallup AI is a Gen AI assistant that uses retrieval-augmented generation (RAG) to provide research-backed answers and recommendations for improving employee engagement and applying CliftonStrengths. It is augmented exclusively with Gallup-owned content to ensure accuracy and does not use client prompts, responses, or other client data to train the AI models, preserving confidentiality.
Survey Reporting Insights (AI)
Employee engagement reporting uses AI to analyze team data and highlight meaningful insights from the team’s collective results. Although AI identifies the insights, the content is grounded in Gallup’s validated engagement metrics, decades of research, observed trends and comparative benchmarks from Gallup’s extensive database.
Survey Reporting Next Steps (AI)
Employee engagement reporting also applies AI to evaluate team results and provide relevant, research-based recommendations tailored to the team’s engagement metrics. These recommendations help managers lead meaningful team discussions, focus on strengths and address opportunities to deliver higher engagement and performance.
Comment Analysis (AI and ML)
Text Analytics on survey reports employs AI to interpret open-ended survey comments with natural language processing (NLP) so it can understand the structure and meaning of written feedback. It then applies ML to classify sentiment (positive, negative, neutral or mixed) and identify common themes and subtopics, helping managers quickly understand key issues and insights from large volumes of qualitative data.
Comment Summaries (Gen AI)
Text Analytics on survey reports also uses Gen AI to create concise summaries of open-ended survey comments by question and topic. These summaries highlight key points from large volumes of unstructured feedback, helping managers quickly process the data and develop clear, actionable insights.
PII and Toxicity Detection (ML)
Integrated into the Text Analytics feature of survey reports, this ML feature flags personally identifiable information (PII) and toxic language in open-ended survey comments. Admins can review and redact flagged responses to maintain data integrity and protect confidentiality.
Comment Translations (ML)
This ML-based feature translates open-ended responses into U.S. English using AWS translation services. It supports over 50 languages, enabling global organizations to review feedback in a consistent format.
Personalized Team and Partnership Insights (Gen AI)
CliftonStrengths features AI-generated insights for teams and partnerships that delivers personalized summaries of key patterns in collaboration, complementary strengths and potential obstacles.
Strengths-Based Insights (AI)
Gallup Access uses AI to analyze team CliftonStrengths data and deliver relevant, research-backed advice based on the team’s collective strengths and potential gaps. While AI helps determine which insights to show, Gallup scientists and subject matter experts write and curate all insights.
Responsible AI Practices
Gallup applies a rigorous framework to create AI systems that are ethical, secure, explainable and privacy-compliant. These practices span all phases of AI development and deployment in Gallup Access, including:
Scientific Integrity and Guardrails
Gallup grounds its AI systems in proprietary research and never trains models on client-specific data. Guardrails such as AWS safety layers, internal rule-based filters and Gallup’s evaluation and configuration of Anthropic Claude model variants, help reduce hallucinations, support consistent output quality and mitigate bias while upholding Gallup’s quality standards.
Compliance With Global Standards
Gallup evaluates all AI implementations using standards such as the NIST AI Risk Management Framework and the OWASP Top 10 for Large Language Model Applications, promoting the development of safe, secure and trustworthy AI.
Transparency and Explainability
Gallup ensures that all Gen AI outputs are traceable to Gallup-owned content (e.g., research articles, guides), making responses auditable, interpretable and aligned with Gallup’s published science.
Privacy and Data Protection
Gallup removes PII before processing with AI tools and houses all AI applications in a secure AWS cloud environment. Gallup never shares data externally or uses client content to train AI models. While Gallup retains AI interaction data (prompts and responses) to support Gallup Access functionality and oversight, it never uses the data to train the AI models and can delete an individual’s AI interaction data at their request.
Ongoing Monitoring and Quality Control
Gallup employs continuous automated and manual testing to ensure reliability, fairness and accuracy. QA teams regularly assess performance, perform bias checks, and respond to client and user feedback.
These practices ensure all of Gallup’s AI tools, features and capabilities are safe, transparent and scientifically sound, while protecting client confidentiality and maintaining trust.