{
  "id": 7764513,
  "title": "The last mile of AI: Closing the gap between insight and action",
  "url": "https://urgent.news/2026/09/16/the-last-mile-of-ai-closing-the-gap-between-insight-and-action",
  "topic": "ai",
  "section": "AI",
  "published": "2026-09-16T10:17:05.000Z",
  "source": {
    "name": "TechRadar",
    "slug": "techradar",
    "url": "https://www.techradar.com/pro/the-last-mile-of-ai-closing-the-gap-between-insight-and-action"
  },
  "original_language": "en",
  "account": "In recent years, the focus has been on AI models and their capabilities. However, organizations are still grappling with how to turn AI-generated insights into actual business outcomes. Despite the progress in AI technology, the real bottleneck lies in connecting the intelligence to how the business operates. AI models can provide recommendations quickly, but acting on them is often challenging. Issues like incomplete data, missing context, and governance concerns can hinder adoption.\n\nFor AI to create value, it needs a clear understanding of business processes, enough context to interpret events, and the ability to function within governance guidelines. The real challenge often lies in the connection between the model and the business process. Data is frequently fragmented across different systems, and essential context is absent. Moreover, teams may define the same concepts differently, making it difficult to create a unified view.\n\nMany organizations find themselves struggling to integrate AI into their existing business systems. Traditional ERP platforms, CRM applications, and operational databases provide the authoritative records of the business. However, AI increasingly requires access to information from outside these core systems, such as supplier data, partner information, and external intelligence sources. Additionally, live operational signals from customer interactions, fraud alerts, and connected devices are becoming crucial for real-time decision-making.\n\nTo make AI effective, organizations must adopt a broader perspective of the business. This involves combining authoritative data within core systems, contextual information from external sources, and real-time operational awareness. For instance, a manufacturer successfully implemented AI-driven fraud detection, but the process slowed down due to scattered customer records, transaction histories, and fraud signals across multiple systems. By bringing these sources together in a trusted, governed view, the organization reduced fraud losses and improved efficiency.\n\nThe rise of AI has also led to a shift in how businesses approach data sharing. Decisions now depend more on information outside organizational boundaries, such as supplier and logistics data for manufacturers, and external signals for banks detecting fraud. Public sector organizations also require information from multiple agencies to enhance services. Trusted AI outcomes depend on secure access to data across distributed environments, creating a balancing act for technology leaders. They must enable data accessibility while maintaining governance, security, and ownership. Without this foundation, even the most advanced AI models cannot deliver consistent results. Ultimately, success in enterprise AI adoption hinges on establishing the necessary conditions for AI to operate effectively, starting with trusted data.",
  "summary": "Discover why AI’s biggest challenge isn’t intelligence, but turning trusted data into measurable business results.",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}