TL;DR: Recent Portswigger research introduced novel HTTP desync techniques discovered through an AI-assisted research system called the HTTP Terminator. The findings expand the range of unusual HTTP ...

The updated protocol will simplify how enterprises deploy and scale MCP-based AI applications in the cloud, but developers relying on deprecated capabilities will need to plan for architectural ...

AI agent protocols IETF standard review begins at IETF 126 Vienna this week, where the agentproto Birds-of-a-Feather session Thursday could charter a working group to produce the first formal RFC for ...

LF Projects is converting the Model Context Protocols (MCP) to completely stateless communication. With the new version, which is scheduled for release today, sessions and initialization handshakes ...

MCP makes models carry session handles, improving scalability while making reliable correlation and authorization an application concern.

IETF 126 in Vienna reaches its most consequential session Thursday as the agentproto BoF votes on chartering a Working Group to produce the first binding RFC for AI agent communication — deciding ...

The company has patched a critical pre-authentication flaw in TeamCity that could let attackers execute arbitrary commands, expose credentials, and compromise supply chains on self-hosted servers.

Cloudflare announced today that it is giving AI agents something the agentic commerce stack has been missing since it began assembling in earnest: a structured, human-controlled budget. The company ...

Coinbase has positioned itself at the center of this trend with the introduction of AgentKit, which allows application ...

Cloudflare Inc. reported today that distributed denial-of-service attacks above 1 terabit per second grew sixfold between the ...

Oracle beefed up its Kubernetes automation provision through its AI database suite and a new Red Hat partnership. The updates are led by Oracle AI Database Operator for Kubernetes v2.2.0, which aims ...

Tracked as CVE-2026-63077, the critical bug can be exploited without authentication for remote code execution.