AI & Machine Learning

A Sinch survey of 2,527 AI decision-makers finds 74% have shut down live AI customer communications agents, with rollback rates rising to 81% among the most governance-mature organizations.

Cloudflare detailed the custom infrastructure behind its Workers AI expansion to host models like Kimi K2.5, including a Rust-based inference engine and a lossless weight-compression system.

Nine days after v0.13 closed Hermes Agent's durability gap, Nous Research shipped v0.14 on May 16 — adding pip install from PyPI, native Windows, and a local OpenAI-compatible proxy that wraps Claude Pro, ChatGPT Pro, and SuperGrok subscriptions.

After an Alpha Omega-mediated Mythos scan turned up five claimed vulnerabilities, the curl security team triaged the list to one low-severity CVE bound for curl 8.21.0 in late June.

A University of Pennsylvania team led by Vivek Shenoy has published Mollifier Layers in TMLR, a module that swaps recursive automatic differentiation for convolutional smoothing to make physics-informed neural networks stable on high-order, noisy inverse PDEs.

After a March 2026 launch that left users contending with broad ACL changes and stale sandbox profiles, OpenAI published a stricter design splitting Codex into offline and online local users with DPAPI-protected credentials and a command-runner handoff.

OpenAI launched a personal finance preview in ChatGPT on Friday for Pro subscribers in the US, using Plaid to pull read-only data from more than 12,000 financial institutions into a single dashboard.

Red Hat used its May 12 Summit keynote to take Red Hat Desktop with Podman Desktop and isolated AI agent sandboxing to general availability, and to preview Red Hat AI 3.4 with vLLM speculative decoding, Model-as-a-Service, MCP tooling, and an agent evaluation hub.

At its May 12 Android Show, Google introduced Gemini Intelligence and a new Googlebook laptop category built with Acer, ASUS, Dell, HP and Lenovo, with Chromebooks able to move to Googlebook-style software via firmware update.

EdgeCore Digital Infrastructure closed $1.5 billion in senior secured term loans for AS01 and AS02 in Sterling, Virginia, two single-tenant facilities totaling 114 MW and 685,000 square feet, with AS01 going live in November 2026 and AS02 in July 2027.

A majority-owned joint venture led by TPG embeds OpenAI engineers inside enterprises to convert frontier model access into measurable business outcomes.

Nvidia receives a five-year option on 30 million IREN shares at $70 each while paying IREN $3.4 billion over five years for Blackwell cloud capacity at Childress, Texas.