OpenAI is putting $1 billion behind cyber defense, according to CoinDesk, by subsidizing access to its Daybreak security platform. The move follows OpenAI’s claim that Astra, an AI system, can independently find previously unknown software flaws and turn them into working attacks, CoinDesk reports. Those previously unknown flaws are commonly described as zero-days because defenders have had no prior notice or patch window. The provided reporting does not specify how the $1 billion commitment is structured, who qualifies for the Daybreak subsidy, or what limits OpenAI places on Astra’s use. It also does not include independent validation of Astra’s zero-day-finding capabilities. That makes this a significant but still developing cybersecurity story. The core claim is not just that OpenAI is funding defensive tooling; it is that the company says its AI can move through the full exploit chain from unknown flaw discovery to a working attack, while also subsidizing a platform positioned for defense. Who benefits: Security teams that can access subsidized Daybreak may benefit if the platform improves detection, testing, or remediation workflows. OpenAI also benefits if it can frame powerful vulnerability-finding AI as a defensive capability. Who's exposed: Software vendors and operators are exposed if AI systems accelerate discovery of unknown flaws faster than organizations can patch. The current reporting does not establish whether Astra’s capabilities have been independently tested.