COMPEL Glossary / artificial-narrow-intelligence-ani
Artificial Narrow Intelligence (ANI)
Artificial Narrow Intelligence describes AI that performs a specific task within a defined domain, such as detecting fraud, translating languages, or playing chess.
What this means in practice
Every AI system deployed in enterprises today is narrow AI -- including the most impressive large language models, which operate within the domain of language processing despite their remarkable breadth. Understanding that all current AI is narrow is strategically important: it means organizations should design programs expecting AI to excel at specific, well-defined tasks rather than replace broad human judgment. Organizations that plan for narrow AI and are pleasantly surprised by broader capabilities will outperform those that plan for general intelligence and are disappointed by reality.
Why it matters
Understanding that all current AI is narrow — including impressive large language models — is strategically critical for setting realistic transformation expectations. Organizations that design programs expecting AI to excel at specific, well-defined tasks consistently outperform those planning for general intelligence. This understanding prevents over-promising to stakeholders and enables focused investment in use cases where narrow AI can deliver measurable, immediate value.
How COMPEL uses it
During the Calibrate stage, organizational expectations about AI capabilities are assessed and calibrated against reality. The Organize stage designs AI literacy programs that build accurate understanding of what narrow AI can and cannot do. The Model stage applies this understanding to use case prioritization, ensuring the transformation roadmap targets specific, well-defined problems where narrow AI excels rather than vague objectives that assume general intelligence.
Related Terms
Other glossary terms mentioned in this entry's definition and context.