AI in Security

OpenAI’s announcement says Daybreak Red is now available in AWS GovCloud (US-East), but provides no product, access, pricing or compliance details.

More than 115 stores across at least six countries reported refrigeration disruptions overnight. The attacker and any role for AI remain unknown.

Binance says an attack drained funds across its wallets on Oct. 29. The exchange estimates a $6.3 billion impact and says about $3 billion was frozen.

NebuSec says answering a Telegram call could trigger an Android exploit chain. Telegram fixed the reported bug in version 12.10.4; key details remain unverified.

Nearly 13,934 CVE records were published by Oct. 25, 2026. The data shows a sharp rise, but does not establish how much AI drove it.

PHP released fixes for two critical remote code execution flaws on Oct. 21, after research by Calif, OpenAI and Wiz traced a related issue back decades.

PHP fixed two critical PHP2Shell flaws and Node.js patched ten vulnerabilities on Oct. 21. Operators still need to check that fixes are installed.

Anthropic’s headline says Claude revealed an ECDSA curve’s origins, but the available material names no curve or evidence for the claim.

Security researchers used Anthropic’s Claude to breach an OpenAI product, raising concerns over AI-enabled cyberattacks. Details are still emerging.

GPT-6 Astra reportedly decrypts a WWI German radio cipher, marking a breakthrough in historical code-breaking and AI capabilities. Details are still emerging.