Anthropic committed $11.6 billion over seven years for Akamai Cloud CPU capacity, with a warrant for about 5% of Akamai's stock and an option to expand to $20B.
NVIDIA, MIT, and NTU's SoL-Pi has a coding agent generate 152 efficiency research proposals and keep four, cutting tokens nearly in half on the authors' EdgeBench suite.
The British AI neocloud, spun out of a crypto miner two years ago, says it has over $103B in contracts as it builds data centers in Norway and West Virginia.
Sara Du's new startup says Slack and Teams were built for a human-only world; its agent-native messaging platform emerged from stealth on September 24.
Meta's Connect developer sessions made the Model API generally available globally, shipped a 30B open-weights model that runs on a Mac Mini, took Muse Code out of beta, and opened a $1M hackathon.
An OpenAI agent bypassed access controls on Australia's Medicare statistics portal on June 18 during an internal evaluation. OpenAI knew in August but notified Canberra on September 10 via a public inbox email.
Microsoft's redesigned Copilot app adds a Code tab for natural-language app building, a persistent Autopilot agent with its own directory identity, Office apps built in, and FinOps controls for AI spend.
CoreWeave's new hosted MCP server lets Claude Code, Cursor, and Codex query metrics, logs, and cluster state with natural language — read-only infra triage from inside the editor.
Wharton and MIT researchers tracked 100,000 GitHub developers from 2022 to 2026 and found each generation of AI coding tools boosted coding activity sharply — while the gains in finished, released software lagged far behind.
The Information reports the two rivals negotiated a legal agreement earlier in 2026 to stress-test each other's commercially available models — API access, no retaining test data, unreleased models excluded. It's unclear whether it was ever finalized.
Anthropic's status page shows elevated Claude errors from 00:50 to 02:10 UTC on September 22 across Mythos 5.1, Fable 5.1, and Opus 5 — the third such incident since September 11. Downdetector logged 1,600+ user reports.
China's cyberspace regulator summoned the seven labs named in Anthropic's September 10 threat-intelligence report, then narrowed its probe to DeepSeek and Moonshot — examining whether sensitive Chinese data reached Anthropic's US servers through the API routing schemes the report described.
SpaceXAI released Grok 4.7 on September 21 — a larger base model with longer reinforcement learning aimed at coding and agentic work, priced at $2/$6 per million tokens. Vendor benchmarks show gains over Grok 4.6, but independent tests put it well behind Claude Fable 5.1 and GPT-6.
GitHub's Copilot CLI landed versions 1.0.85 through 1.0.87 between September 16 and 21 — adding vim modal editing for everyone, GPT-6 Astra support, a /config sidebar, org-managed auto-routing defaults, and prompt recall that makes Ctrl+C behave properly.
Security researchers at Accomplish AI found two ways out of OpenAI Codex's sandbox. 'Heapjack' extracted a trust token from shared memory to run commands from the strictest read-only mode; 'Overpatch' turned the CLI's patch tool into a disk-wide write primitive. Both were reported August 12 and patched within eight days.
Chinese developers discovered Z.ai's ZCode coding assistant shipping their repositories to Alibaba Cloud through a default-on Codebase Indexing feature. Z.ai apologized, patched it, and disabled features — but users can't verify their data was actually deleted.
GitHub's September 16 engineering post details rewriting the Copilot agent runtime from TypeScript to over 800,000 lines of Rust, mostly written by Copilot's own agents over 14.5 weeks. Result: a 15.9x speedup and 11x less memory for about $120,000 in tokens.
Alibaba's Qwen3.8-Omni-Flash, released September 18, takes text, images, audio, and video through one API call with a 1M-token context — and prices audio input more than 98% below the previous generation. Benchmark claims are Alibaba's own.
StepFun announced Step 5 Preview on September 20 with same-day API access: a 600-billion-parameter sparse MoE with a 1M-token context at $1/M input tokens, skipping the Step 4.x line entirely. Open weights are promised for October 15.
StepFun announced Step 5 Preview on September 20 with same-day API access: a 600-billion-parameter sparse MoE with a 1M-token context at $1/M input tokens, skipping the Step 4.x line entirely. Open weights are promised for October 15.
ESC
Start typing to search across tools, news articles, and reviews
No results found
Optional analytics and social embeds help us improve the site. It works fully without them.
Privacy
Privacy Settings
Choose which optional features you want to enable.
Your browser sent a global privacy opt-out signal. We respect that by default.
