BARC AI and Generative AI Policy Statement

Introduction

BARC has analyzed and covered artificial intelligence (AI) since the 1990s. We now employ generative AI and large-language models (LLMs) to support our research, events, and consulting work.

Analysts (not machines) remain responsible for every insight we publish.

Guiding principles

• Transparency – We disclose how AI supports our work.
• Human value-add – Human insights, expert judgment, context, and synthesis drive every deliverable and monitor accuracy.
• Accuracy and relevance – All outputs meet BARC editorial standards before publication.

Permitted AI uses

• Analyst content: Drafting outlines, titles, summaries, and first-pass text that analysts refine.
• Checking grammar, reformatting copy, and summarizing long transcripts.
• Optimizing human originated content
• Rapid public-source research, thoroughly fact-checked by analysts.
• Tagging and translation tasks reviewed by subject-matter experts.
• Creating images for presentations, blogs, and social posts.
• Marketing use cases such as blogs and social posts derived from published content.
• Analyzing data sets and data visualization
• Ideation

Prohibited or restricted AI uses

• Stand-alone analysis or recommendations
• Synthetic interviews and synthetic responses on BARC surveys.
• Automated replies to client questions.
• Input of client-confidential data or unpublished research into external AI tools.
• Any output offered “as is” without analyst review.

Value assurance

Every BARC deliverable combines proprietary research and human expertise. AI assists but never replaces our analysts. Clients receive the depth, context, thought leadership, and independence that define our work.

Data protection

BARC applies strict data governance. We keep personal data, client-confidential material, and gated research out of external AI systems. Internal AI tools safeguard privacy, sovereignty, and intellectual property.

Continuous review

We update this policy as AI capabilities, regulations, and client needs evolve. Questions or feedback: [email protected].