{
  "id": 14011148,
  "title": "WindBorne wins a hefty US Air Force contract for its balloon data",
  "url": "https://urgent.news/2026/10/12/windborne-wins-a-hefty-us-air-force-contract-for-its-balloon-data",
  "topic": "tech",
  "section": "Tech",
  "published": "2026-10-12T17:19:14.000Z",
  "source": {
    "name": "TechCrunch",
    "slug": "techcrunch",
    "url": "https://techcrunch.com/2026/10/12/windborne-wins-a-hefty-us-air-force-contract-for-its-balloon-data/"
  },
  "original_language": "en",
  "account": "The U.S. Air Force has awarded WindBorne Systems, a Redwood City, California-based startup, a substantial contract that could be worth nearly $50 million. The contract focuses on purchasing data collected by the company's fleet of autonomous weather balloons, which currently number around 700 spread across the globe. These balloons gather atmospheric sensor readings, which WindBorne sells to customers and feeds into its AI weather forecasting model.\n\nFounded in 2015, WindBorne is one of the pioneering organizations to directly integrate empirical data into a deep-learning weather model. The company has established close ties with the U.S. government, particularly the National Weather Service, the U.S. Air Force, and the U.S. Navy. However, this is the largest contract WindBorne has secured to date.\n\nThe contract is worth up to $49 million, but the Air Force does not pay the full amount upfront. Instead, they pay as they acquire the needed data, placing orders for specific regions and times. So far, WindBorne has received $1.7 million in payments, and executives anticipate additional orders in the future.\n\nThe contract serves as a means to support the growth of WindBorne's balloon constellation and to further research and development efforts aimed at enhancing the hardware and performance of the balloons. WindBorne's co-founder and CTO, Andrey Sushko, told TechCrunch that the contract represents a strong affirmation of the company's ability to provide the required capability for global data collection.\n\nWindBorne CEO John Dean revealed in August that the company had raised a $37 million Series B round to continue scaling its business. Earlier this summer, the company unveiled the latest version of its forecasting model, WeatherMesh 6, which claims to deliver more accurate forecasts than the traditional model employed by the European Centre for Medium-Range Weather Forecasts (ECMWF). The ECMWF simulates atmospheric conditions using the laws of physics.\n\nWith the emergence of transformer model architectures and the release of a vast weather dataset by the ECMWF, AI weather forecasting has become a highly productive application of deep-learning technology. Meteorologists previously relied on costly supercomputers running intricate equations to approximate atmospheric behavior. However, with the advent of end-to-end learning, AI models can ingest decades of weather data and generate predictions more swiftly.\n\nBoth frontier labs and government agencies are developing AI weather forecasting models that demonstrate greater accuracy than traditional forecasts. Meteorologists still require a combination of traditional and AI models to achieve a comprehensive understanding of weather patterns. In August, WindBorne made history by becoming the first organization to dispatch an unpowered flight system—without an engine—to gather data in the eye of a typhoon. The company navigates its balloons by adjusting their altitude, bleeding air, or releasing sand as ballast to reach the desired locations. This novel step is made possible by the availability of robust wind and weather models, WindBorne's large balloon constellation, and the company's ability to operate in the right areas at the appropriate times.",
  "summary": "A Silicon Valley startup that uses weather balloons and AI to predict the weather won a US military contract worth almost $50 million.",
  "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."
}