Two security vendors put OpenAI's cyber models into exploit prioritisation in two days
SentinelOne and Sophos have both put OpenAI's cyber models into exposure prioritisation services within two days. Neither product is generally available and neither company disclosed pricing.
SentinelOne has expanded its Wayfinder Frontier AI Services using OpenAI's Daybreak models, starting with GPT-5.6-Cyber, adding code risk analysis, compromise assessment and malware analysis with findings reviewed by its own analysts. It follows Sophos, which announced on 7 September that it is building an Exploit Path Verification capability into Sophos Managed Risk using the same OpenAI models through the same Daybreak Defense Network. Both frame the problem identically, as vulnerability backlogs that do not tell defenders which exposures an attacker can actually reach. Neither is generally available: SentinelOne's is in private preview and Sophos says availability will be announced later. Neither company disclosed pricing, India availability or any customer.
