Anthropic sets IPO price range of $70 to $78 per share to reach an implied valuation of $1.404 trillion

Anthropic set an IPO price range of $70–$78 per share for 825 million Class A shares, saying a top-of-range listing could imply a $1.404 trillion valuation. The Claude maker also agreed to privately sell $26 billion of non-voting stock to AMD, Broadcom, Google, Micron, Microsoft, Nvidia and TSMC, with Google and Nvidia each investing $10 billion.

How the Next Congress Can Repair the Military’s Personnel Crisis

The article argues the next Congress should use the fiscal 2027 defense authorization bill to repair military personnel policy, citing a survey in which 53 percent of active-duty respondents said they were considering separating early. It calls for protecting vetted officers from removal from promotion lists without notice or appeal and for barring gender-based exclusion from combat roles.

Minneapolis woman dragged from car during Metro Surge sues feds for assault, imprisonment

Aliya Rahman, a 44-year-old south Minneapolis resident, sued the federal government alleging assault, battery, false imprisonment and negligence after agents dragged her from her car during a Jan. 13 immigration enforcement operation. She filed the civil suit after the government did not respond to her April tort claim within six months.

Blind Eagle Uses Fake Court Notices to Deliver Remote Access Trojan Across Latin America

Blind Eagle, also tracked as APT-C-36, has launched a new Latin America-focused phishing campaign that uses Spanish and Portuguese judicial and administrative lures to deliver a remote access trojan. Zenox reported that the campaign observed between late October and early November 2026 involved nearly 3,300 malicious emails, though it could not confirm any successful real-world compromises.

Hackers exploit trusted AWS access to breach multiple cloud environments

A threat actor used compromised IAM credentials for Qubole's AWS account to assume roles in AWS environments of tens of companies between October 14 and 26, Wiz researchers found. The attacker retrieved secrets and downloaded data from accessible S3 buckets, and Sentra said dozens of victims were affected, with millions of files stolen from one customer.

iOS Sandbox Escape Is Public, and Apple Hasn’t Confirmed a Fix

Security researcher forcequitOS has published public proof-of-concept code for an iOS sandbox escape called worse_query, which Binance says could expose seed phrases, private keys and login credentials. Binance and the researcher have not reported any attacks or data theft, and Apple's updates did not mention a fix.

OpenAI lets businesses claim ChatGPT profiles and launch their own agents

OpenAI began a beta called Businesses in ChatGPT letting companies claim verified profiles, update prices and stock, and see aggregated mention and click data; eligible US ecommerce and online-service firms can also launch Business Agents that chat with customers. Those agents can also be sold as ad placements, must state they are AI, and OpenAI's advertising reached a $1B annual run rate in August.

Russia-Linked APT Turla Turned Exchange Servers Into Their Own Control Room

ESET says it found a previously undocumented Turla toolset named Cereal on government Exchange servers in Kazakhstan and Romania, attributing the toolset to the Russia-linked group with high confidence. In Romania it was set to spot NATO and EU email traffic, and a passive backdoor takes commands via Exchange web paths rather than an external server.