Full-text search across 69574 articles. Combine with topic and date filters; results sorted by relevance.
Pavel Rabtsevich reported on September 23rd that his Jev model matched NASA Exoplanet Archive dispositions on 5,840 of 8,054 Kepler Objects of Interest, a 72.51% match rate, after a prompt rewrite that followed a first r…
Oracle Corp. is shifting agent security controls to the data layer, enforcing authorization at the row, column and cell level through Oracle Deep Data Security based on user identity and runtime context, according to Jua…
Cedar AI released the beta of AOS GLM, a C++20 deterministic constrained language engine that architect Joshua Edward Osborne says runs entirely in host memory with 0.000 MB GPU VRAM and sub-microsecond decision latencie…
Anthropic announced that its Claude AI agents autonomously discovered a novel enzyme system, which the company calls array-associated reverse transcriptases (ART), found in bacteriophages and associated with an array of …
ServiceNow, Rubrik, Microsoft and Lumos shipped MCP-layer agent governance products within a 12-day span in September, converging on enforcing policy at the protocol layer rather than in the model or application. Service…
A blogger tested Claude Opus 5.5 by tasking the model with reviewing three prior blog posts about a browser-based web aquarium and rebuilding it on a feature branch as an extensible aquarium, with no new assets or packag…
Devframe launched a framework-neutral foundation that lets developers build a devtool once and run it inside any host framework, as a standalone tool, or through a coding agent. The project offers a getting-started wizar…
Anthropic's Opus 5.5 and OpenAI's GPT-6 Sol both cut API prices and cost roughly half as much per coding task as the models they replace, according to a 12-task benchmark run three times each on all four models. Opus 5.5…
GPT-6 Astra, run through Codex at medium reasoning effort, completed 100% of the DrivingBench cone course in 5:22, the only model to finish among four frontier models given control of a Toyota Corolla's steering, acceler…
Daytona cofounder and CEO Ivan Burazin said the agents he uses fail more often because they cannot access the tools they need than because of the model itself, and that he does not want agents operating under his own cre…
Aikido Security has partnered with Deel to bundle its security platform into Deel's global hiring ecosystem, covering code, devices and cloud access for every worker a company onboards through Deel. PropertyGuru is one o…
Mastra, an open-source TypeScript framework for AI applications and agents, has been updated as of September 2026 and now holds 28,000 stars on GitHub. The framework routes to more than 40 model providers through a singl…
Zed, the Rust-based code editor from the creators of Atom and Tree-sitter, has 91,000 GitHub stars and was last updated in September 2026, according to its public listing. The editor ships native agentic editing, support…
Google's open-source Agent Development Kit for Go (Google ADK) has been updated, with the code-first Go toolkit for building, evaluating and deploying AI agents now at 8.8k GitHub stars as of its September 2026 release. …
OpenAI launched GPT-6 Sol and Luna on September 22, 2026, with the company claiming the models ship at half the API cost of the GPT-5.6 series. OpenAI says Sol cuts factual errors roughly in half, reaching Astra-tier rel…
Microsoft's Digital Crimes Unit, working with Cloudflare, Coinbase, OpenAI, Railway, SpyCloud, Shadowserver Foundation and TRM Labs, disrupted the AI-powered phishing-as-a-service platform EvilTokens, which had compromis…
In the sixth essay of his AI Harness Engineering series, developer Derek Wang argues that resilient implementation — idempotency, graceful degradation, visibility, and fast recovery — is the foundational layer of any AI-…
A developer argues that AI agents need dedicated payment infrastructure, comparing three custody models: agents holding private keys (fast but exploitable), multi-sig with human approvals (secure but too slow for microtr…
Google DeepMind's 2020 catalogue of around 60 specification-gaming examples, including a boat-racing agent that circled and hit the same reward blocks instead of finishing the race, is the basis for a new argument that b…
A threat actor published malicious releases of MemTensor's MemOS AI memory framework on September 23, 2026, compromising npm packages @memtensor/memos-cloud-openclaw-plugin versions 0.1.21, 0.1.23 and 0.1.25 and PyPI pac…