{
  "id": 9005223,
  "title": "The Best Enterprise AI System Is One CFOs Are Allowed to Use",
  "url": "https://urgent.news/2026/09/21/the-best-enterprise-ai-system-is-one-cfos-are-allowed-to-use",
  "topic": "ai",
  "section": "AI",
  "published": "2026-09-21T20:52:57.000Z",
  "source": {
    "name": "PYMNTS",
    "slug": "pymnts",
    "url": "https://www.pymnts.com/news/artificial-intelligence/2026/the-best-enterprise-ai-system-is-one-cfos-are-allowed-to-use/"
  },
  "original_language": "en",
  "account": "The most critical factor in the future of enterprise artificial intelligence isn't the model's intelligence, but its data retention policies, according to recent developments. If the recent news from companies like Nvidia, Palantir, Booz Allen Hamilton, and Novo Nordisk are any indication, the importance of data retention policies is becoming increasingly clear. As AI becomes more useful within an enterprise, the sensitivity of its context grows, which in turn alters the economics of model selection. A slightly better reasoning score may be valuable, but it's not worth it if accessing that capability requires a company to alter its longstanding rules governing confidential information. For CFOs, CIOs, and CISOs, the question shifts from \"Which model performs best?\" to \"Which model can perform best within the controls we already have?\"\n\nIn the early years of generative AI, companies compared models based on reasoning, accuracy, latency, and price. However, as AI moves out of experimentation and into finance, cybersecurity, legal, and other information-rich functions, data retention policies are becoming a more prominent factor in procurement decisions. Companies like OpenAI, Google Cloud, and Anthropic are offering various approaches to data retention, such as zero data retention (ZDR), private safety processing, and Enterprise Frontier Safeguards (EFS). These policies allow companies to keep data on their own infrastructure, under their own encryption keys, while still benefiting from the AI model's capabilities. This new competitive front among AI providers isn't just about building the smartest model, but delivering frontier intelligence without requiring enterprises to surrender control of their underlying data.",
  "summary": "The most important benchmark for the next phase of enterprise artificial intelligence isn’t arising around the model’s intelligence. Instead, the most important benchmark is around a model’s data retention policies. That is, if the news last week that companies as varied as Nvidia, Palantir, Booz Allen Hamilton and Novo Nordisk are drawing hard boundaries around […] The post The Best Enterprise…",
  "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."
}