Stories about Microsoft Azure
1 related stories
OpenAI says it stopped a campaign to steal its models' reasoning, but the trick still worked on Azure
AI InsightAlthough OpenAI blocked over 15,000 accounts targeting model reasoning, the attack remained effective on Azure (including GPT-6 Astra) for weeks. This exposes a mechanism gap where core security protections failed to sync into third-party cloud ecosystems during API resale.Key TakeawayOpenAI's direct protection fails to cover cloud platforms, exposing a security responsibility disconnect in API reselling.Why It MattersIt reflects a power-and-responsibility vacuum in security policy synchronization between tech providers and cloud vendors in multi-cloud distribution, directly threatening asset security.Who's Affected- At RiskOpenAICore model reasoning assets being continuously stolen on third-party clouds directly exposes blind spots in ecosystem security.
- At RiskMicrosoft AzureHosted models suffered unblocked security breaches for weeks, facing scrutiny over risk-control capabilities.
What's NextMonitor whether OpenAI will tighten security baselines for multi-cloud distribution and whether cloud vendors will upgrade real-time defense for hosted APIs.Importance 60/100