Sep 15, 2026ToolsResearchToolsBairesDev Survey Finds 42% of Developers Say AI Writes at Least Half Their CodeThe new survey suggests that faster code generation is not shrinking engineering work. It is redirecting time and organizational spending toward validation, debugging and learning.2 min
Sep 15, 2026ToolsLaunchToolsTemplafy Connects Claude, ChatGPT and Copilot to Company Document RulesThe new connectors let employees start in a preferred AI assistant, then turn that work into editable documents using company-approved materials and requirements.2 min
Sep 14, 2026ToolsToolsAnthropic Rebuilds Its Test System After AI Coding Drives 25-Fold CI GrowthThe company says its distributed replacement has stabilized a growing backlog, but the scalable design costs more to operate—an early operational trade-off of agent-driven software output.4 min
Sep 14, 2026ToolsResearchToolsMicrosoft Research Finds Copilot Retries Can Quadruple Compute CostsA production-scale snapshot of GitHub Copilot shows why tool reliability and context retention matter to the cost of autonomous coding work.2 min
Sep 14, 2026ToolsResearchToolsLlamaIndex Publishes Two-Pass OCR Workflow for Small Document Data RoomsThe proposed workflow scans every uploaded file cheaply, then reserves more expensive visual parsing for pages an agent identifies as relevant.2 min
Sep 13, 2026ToolsOpen releaseToolsLangChain Open-Sources Slack Ad Agent With Human Approval for ChangesThe tool links ad-platform activity to pipeline outcomes, but its design puts calculations in code and campaign decisions behind a human approval step.2 min
Sep 12, 2026ToolsToolsOpenAI Profiles Lab Using ChatGPT and Codex to Speed Antimicrobial SearchCésar de la Fuente’s lab says its AI systems can shrink the first candidate search from years to hours. The harder work—showing a molecule is safe, effective and manufacturable—still happens in the lab.2 min
Sep 12, 2026ToolsEnterprise adoptionToolsOpenAI Publishes a Leaner Prompting Playbook for Codex AgentsThe company’s new guidance argues that accumulated instructions can waste an agent’s context and make it stop at the wrong moment. The harder task is trimming routine guardrails without weakening protections around consequential work.3 min
Sep 12, 2026ToolsOpen releaseToolsGoogle Open-Sources ARTEMIS for Android Automation, Claims 99%+ Benchmark ResultThe tool lets coding assistants operate phones and emulators from plain-English requests. Its two modes make developers choose between quicker execution and more thorough checks.2 min
Sep 11, 2026ToolsSecurity riskToolsResearchers Say OpenAI Agents Used RubyGems Packages to Run CodeThe researchers say the agents used a documentation service to run code and retrieve public data. OpenAI confirmed the episode but described the work as benign internet access while its investigation continues.4 min
Sep 11, 2026ToolsLaunchToolsOpenClaw Adds a Way to Turn Past Chats Into Reusable SkillsVersion 2026.9.4 pairs a steerable chat-to-skill workflow with setup and interface fixes, aiming to make an agent system easier to reuse after the first conversation.3 min
Sep 10, 2026ToolsResearchToolsGoogle Releases ToolGrad, Reporting a 99.8% Pass Rate for Tool-Use DataThe research framework reverses a common synthetic-data workflow: it validates API calls first, then writes the request a user might make. Google says the approach produces more usable examples at lower cost, though its results remain tied to research benchmarks.4 min
Sep 10, 2026ToolsLaunchToolsGoogle Adds AI-Guided Measurement Tools and Opens Meridian GeoX GloballyThe release links customer-data management, conversion measurement and causal testing more closely, while leaving advertisers to judge whether Google’s reported performance averages fit their own setups.3 min
Sep 10, 2026ToolsLaunchToolsOracle Adds a Gateway to Control AI Agents’ Access to Company ToolsPlanned for Oracle Integration 26.10, the new layer is designed to keep credentials, tool permissions and policy checks outside individual agent implementations.3 min
Sep 10, 2026ToolsPlatform shiftToolsPerplexity Will Retire Sonar API Endpoints on September 27, Forcing a MoveApplications using Sonar’s chat-completions endpoint must shift to a different request interface that pairs web search with multi-step research and tool use.2 min
Sep 10, 2026ToolsToolsLangChain’s Beta LLM Gateway Blocks Models Outside Access PoliciesOrganizations can set provider and model limits by organization, workspace, user, or API key. But a narrower rule replaces a broader catalog instead of adding to it.2 min
Sep 10, 2026ToolsToolsPerplexity Publishes an AI Incident-Response Recipe With Read-Only GuardrailsThe new guide combines saved agent settings with live service connections, while placing the most consequential safety boundary in instructions and account permissions rather than the connector itself.3 min
Sep 10, 2026ToolsLaunchToolsClearview AI Built an AI Prototype for Police Profile SearchesInquiryIQ is designed to follow a facial-recognition lead across the web and compile possible personal details. Clearview says it has never been used by a law-enforcement customer and is not planned for release in its current form.2 min