{
  "id": 14434742,
  "title": "Vast Data bets stitched-together AI pipelines can’t keep sensitive data in sync",
  "url": "https://urgent.news/2026/10/14/vast-data-bets-stitched-together-ai-pipelines-cant-keep-sensitive",
  "topic": "ai",
  "section": "AI",
  "published": "2026-10-14T11:15:06.000Z",
  "source": {
    "name": "SiliconANGLE",
    "slug": "siliconangle",
    "url": "https://siliconangle.com/2026/10/14/data-sovereignty-strains-best-breed-ai-pipelines-aidatapipelineforum/"
  },
  "original_language": "en",
  "account": "Data sovereignty is emerging as an architectural challenge as businesses bring artificial intelligence inference closer to their most sensitive information. Pipelines composed of separate tools struggle to maintain alignment between embeddings, permissions, and source data, which hinders many valuable data sets from being utilized in AI. The transition from model training to inference is altering the requirements AI infrastructure must meet, driving demand for Vast Data Inc., which raised approximately $1 billion in April 2026 at a $30 billion valuation. Vaughn Stewart, vice president of systems engineering at Vast, highlighted that enterprises are increasingly required to manage user identification, access control, and regulated data alongside scale and speed. Stewart explained that while some companies adopt a bespoke model approach, Vast's solution focuses on integrated data platforms for AI factories. He emphasized that the architecture combines storage systems and data-processing tools, enabling performance that surpasses traditional methods. Vast's disaggregated, shared-everything architecture streamlines the functions of storage and processing, improving efficiency. Regarding competition, Stewart criticized a reliance on assembling pipelines with swappable components, arguing that moving data between systems and losing access controls can lead to data not being in sync. This approach is seen as a fallacy when serving parallel requests at mass scale. Half of the world's data remains in corporate data centers or colocation facilities due to sensitivity and regulatory constraints, making data sovereignty a significant barrier. Vast's solution to this issue is DataEnclave, a confidential AI runtime announced in September 2026 with Nvidia and Cohere as partners. DataEnclave operates models in secured virtual machines, demands remote cryptographic attestation before granting access to keys, and logs all access in a tamper-proof database. Stewart stated that without attestation, there's no access to keys or data, positioning Vast to bring frontier models to highly secure datasets and bridge the last mile in expanding AI tools.",
  "summary": "Data sovereignty is becoming an architectural problem as enterprises push artificial intelligence inference toward their most sensitive information. Pipelines assembled from separate tools struggle to keep embeddings, permissions and source data aligned, leaving many valuable data sets out of AI’s reach. The shift from model training to inference is changing what AI infrastructure has to […] The…",
  "key_points": [
    "Vast Data raises $1 billion in April 2026 for AI infrastructure",
    "Disaggregated, shared-everything architecture streamlines storage and processing",
    "DataEnclave provides confidential AI runtime with access controls"
  ],
  "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."
}