Find published research by company, product, platform or technology.
Showing 201–220 of 531 dossiers
Stripe says Revenue Recognition users covered by its pricing transition must select a subscription plan by August 19, 2026. If they have not switched by August 20, Stripe will automatically turn the product off until they subscribe.
SwarmLLM does not route whole prompts to separate machines; it pipelines one model across browser tabs. A MacBook and iPhone can jointly hold Qwen 3.8 27B even when neither device can hold the full 15GB quantized model alone, with no inference server in the loop.
The new processor can vary sample rates by trace fingerprint and target either a traffic percentage or throughput budget. It is usable now in Honeycomb’s Collector distribution, while the upstream OpenTelemetry component is still working toward alpha.
Cursor has become a concrete example of coding-tool supplier risk: a corporate acquisition can trigger a frontier-model provider’s change-of-control rights and remove a major model family from the product even when the coding tool itself remains operational.
The pilot attacks a persistent evaluation trade-off: labs do not want to reveal frontier-model internals, while evaluators do not want benchmark prompts leaking back to the model provider. DeepMind says a Singapore AI Safety Institute pilot kept both sides’ sensitive assets hidden during execution.
Tor Browser 16.0 makes the annual Firefox ESR migration real: Linux ARM64 reaches stable, 32-bit Linux support ends and privacy hardening comes with explicit accessibility trade-offs.
The October 15 Rust 1.99.1 release fixes a separate FFI bool miscompilation after September's 1.98.1 vtable bug. Check toolchain pins and rebuild artifacts produced by either affected compiler.
WordPress 7.1.4 is a new security release; 7.1.3 is no longer the latest patched 7.1 build. Operators should update promptly.
GitHub's 2026 platform data shows agent-started PRs rising sharply. The implications concern code review, security and multi-model workflows, not proof of higher software quality.
Node 26 is now production LTS. Teams using floating LTS selectors should test which major their CI, containers and native modules actually run.
GitHub's gh-aw reaches GA, making Markdown-defined agent jobs a supported Actions workflow pattern. Security boundaries, model costs and privileged triggers still matter.
Six of eight SVG sanitizers in a PortSwigger test had bypasses; developers must review SVG upload and rendering contexts, CSP and library versions.
Copilot now offers model-reviewed action approvals between manual and Allow All. Unlike OS sandboxing, this probabilistic gate can approve unsafe operations; unattended CLI flags may leave unexpectedly broad permissions.
Cloudflare discloses cross-tenant GLM-5.3 Workers AI responses caused by a KV-cache rejection bug; SGLang v0.5.22 is recommended for disaggregated inference operators.
Operators of FrankenPHP 1.13.x should upgrade promptly: 1.13.4 closes seven server flaws, including a worker-mode denial of service and request-isolation leaks.
OpenBSD 8.0 operators should apply the first signed errata or syspatch updates promptly, especially where IPsec AES-GMAC and NSD are used.
The useful shift is architectural: agent permissions no longer have to depend only on the model or harness behaving correctly. OpenShell puts policy enforcement in the execution environment, while Sentry is designed to keep watching from a separate hardware trust domain.
With no planned installation banner, apps should identify affected merchants and handle the migration proactively; Shopify's alternative theme app embeds still require activation on the published theme.
Stripe Radar Standard, Plus and Pro merchants need to recheck renewal-screening settings and invoice volume. Radar Lite remains included with standard processing, and only transactions actually screened incur paid-tier screening fees.
Small SaaS teams can now move supported Paddle setup and billing tasks into Claude Code, Codex and similar agent workflows, but production account writes still require careful access control; native metered billing is not generally available.