The AI Recommendation Audit 2026: Which B2B/GTM Software AI Assistants Recommend

An eight-engine audit of which B2B/GTM software AI assistants recommend across buying categories, released in this version as the AI Disagreement Index. Version 2.0 expands coverage from five to eight live AI models (ChatGPT, Perplexity, Gemini grounded, Llama 4, GPT-OSS, Qwen, Groq-served Llama, and Cohere), adds cross-engine disagreement statistics (Fleiss' kappa 0.41, mean pairwise agreement 0.44, and intra-model consistency), and applies tool-name normalization across engines so the same product is scored consistently. Captures reflect an honest 2026-06-19 to 2026-07-08 collection window. Includes per-tool AI-Visibility Scores, per-category leaderboards, cross-engine agreement, and source-citation tallies, with web-grounded and model-memory engines compared directly. Methodology and full sourcing at https://lucreya.com/research/ai-recommendation-audit-2026/ . License CC-BY 4.0.