The Wire
Curated cybersecurity and tech news — AI-summarized, source attributed.
Bend: a proof-checked language that aims to make AI coding bugs unmergeable
Bend is a programming language pitched squarely at AI-generated code. Its central idea is a lightweight formal-verification workflow: developers write invariant
Bonsai 2 27B: a 27B model in 5.9GB that keeps 98% of its benchmarks
PrismML has released Ternary Bonsai 2 27B, a heavily quantized version of Qwen3.8 27B that stores each weight as a ternary value (−1, 0, +1) with FP16 group-wis
Cactus Needle 3: 8–29MB on-device models that punch above their weight at tool calls
Cactus Compute has released Needle 3, a foundation model aimed squarely at edge hardware — phones, wearables, robots, smart-home devices, cars, and even microco
Cloudflare Quick Tunnels add JSON output, courting the coding-agent crowd
Cloudflare's Quick Tunnels expose a local development server to the public internet with a single command—`cloudflared tunnel --url http://localhost:8000`—and n
CrowdSec Confirms Private Source Code Leak Traced to Tanstack Supply-Chain Backdoor
CrowdSec has confirmed that its private GitHub source code was exposed in a May 2026 breach, disclosed to the company on September 16. The exposed material cove
Jemalloc 5.4.0 lands with 160+ commits: pinned-memory tracking, tcache cleanup, deadlock fixes
Jemalloc, the widely deployed general-purpose memory allocator, shipped version 5.4.0 on September 17, 2024, folding in more than 160 commits. The release leans
Laser Fault Injection Beats RP2350's 'Permanent' Debug Lock, Extracts OTP Secret
Researchers at Ledger Donjon defeated the hardware security on Raspberry Pi's RP2350 microcontroller, restoring debug access on a chip whose debug interface had
libheif overflow + SSO flaw let researchers reach OpenAI's internal repos
Researchers at Hacktron chained two bugs to hijack OpenAI employees' ChatGPT and Codex accounts, and the whole path from discovery to internal-repo access took
North Korean nuclear test triggered years of aftershock earthquakes
An underground North Korean nuclear test set off a sequence of earthquakes that continued for years after the detonation. That is the full extent of what can be
OpenJev: browser-only demo pits reading a model's logits against generating JSON
OpenJev is a client-side experiment that runs small open-weight language models entirely in the browser—no backend, no waitlist—to demonstrate two ways of extra
Unsealed NYT filing: Microsoft, OpenAI staff called AI scraping 'theft'
Newly unredacted filings in The New York Times' three-year-old copyright suit against OpenAI and Microsoft quote the companies' own employees describing their A
Unverifiable: no retrievable content for "Astra for Law"
No summary was produced. The submitted item contained only a headline ("Astra for Law") with no article body, and the source URL (openai.com/index/astra-for-law