Overnight AI Brief · October 17, 2026

AI-writtenWritten automatically by AI (Claude) from the linked news stories, without human review. How it works

Replit Drops Anthropic, AI Models Used in Security Tests, and an AI CSAM Arrest

Here's a quick look at three separate AI stories from the past few hours, covering a business split, a security demonstration, and a criminal case.

The takeaway: Today's stories show AI's reach into business decisions, security testing, and serious crime, each raising different concerns about how these tools are used.

3 stories 2 outlets 60 sec read

arstechnica.com 2
techcrunch.com 1

In this briefOpenAIAnthropic

Replit Ends Partnership With Anthropic

Coding platform Replit has stopped working with Anthropic. Replit's CEO said cost was one reason for the split. Replit still works with OpenAI, a partnership it started in August.

Why it matters & sources

Why it matters: It shows how AI companies can quickly change partners based on cost, even after announcing high-profile deals.

Sources: techcrunch.com

OpenAI Models Used to Hack Wallet and Websites in Research Test

In a security research demonstration, OpenAI's AI models successfully broke into a cryptocurrency wallet. The same models also breached several other websites during the test.

Why it matters & sources

Why it matters: This shows AI tools can be used to find and exploit security weaknesses, which raises questions about how these systems could be misused.

Sources: arstechnica.com

New Jersey Prosecutor Arrested Over AI-Generated Child Abuse Images

A deputy attorney general in New Jersey was arrested. Police say he used the Discord name 'thegentlemanpervert' and made AI-generated child sexual abuse material.

Why it matters & sources

Why it matters: The case highlights how AI image tools are being used to create illegal content, including by people in positions of legal authority.

Sources: arstechnica.com

Written with the help of AI (Claude) from the AI news stories our feed collected in that edition’s time window. Summaries are ours and every item links to its original sources — AI can make mistakes, so check the source before relying on a detail. How the brief is made.