AI Agents × Developer Tools

Cohere, OpenText pitch agents to gov

Cohere, OpenText pitch agents to gov

✎ Story body

Whether software that runs office work on your behalf can move from trial into production turns on something other than how clever it is.

What happened

Cohere and OpenText announced a partnership. The goal is to move agents — software that takes an instruction, works out its own steps, and operates one internal system after another — from pilots into production inside governments and heavily regulated industries. Cohere's enterprise models sit on OpenText's corporate data layer so the software can read internal records and work across several systems. Deployment can be chosen: a company's own hardware, or a cloud kept within national borders. That choice is the condition for handling data that cannot leave the building. NVIDIA, separately, released a way to hand an agent the job of rewriting GPU routines in Rust. The translation runs as a pipeline with machine checks at every stage, and NVIDIA says it moved all 24 published operators across.

Why it matters

The two answers are not the same one. What Cohere moved is the boundary — what the software is shown, and where it is allowed to sit. What NVIDIA moved is the stage after: the output is filtered by machine before a person reads it, and only clears once its speed has been measured too. Decide the scope up front, or check everything that comes out the back. Putting this software where mistakes are not tolerated, what each built first was not intelligence but one of those two enclosures. That said, the Cohere announcement is a framework: which jobs actually went live is not stated.

What to watch

Which task the Cohere–OpenText pairing puts into production first. For NVIDIA, whether the same checks hold outside the published operator set — on the idiosyncratic code sitting inside real codebases. That is where an enclosure either proves out or stays on paper.

▲ Official & Press
Official

Cohere and OpenText partner to bring trusted agentic AI to governments and regulated industries

Cohere Blog ・ 2026-09-16 ・ 📌

Cohere and OpenText partner on agentic AI for regulated industries

Official

TensorRT Edge-LLM Completes the MLPerf Edge Agentic Benchmark 6.4x Faster on Jetson AGX Thor

NVIDIA Developer Blog ・ 2026-09-16

NVIDIA runs MLPerf Edge Agentic 6.4x faster on one Jetson AGX Thor

Community

Claude Cowork and chat are now one Claude

Simon Willison's Weblog ・ 2026-09-16

Anthropic merges Claude Cowork and chat into one Claude

Official

Translating CUDA Tile Operations from Python to Rust Using Agentic AI

NVIDIA Developer Blog ・ 2026-09-16

NVIDIA uses an AI agent skill to port TileGym kernels to cuTile Rust

Press

Devinが仮想環境でmacOSの提供開始。Macの実機不要でDevinがコード生成、テスト、デバッグ、実行、AppStore配信前のベータ公開まで実行

Publickey ・ 2026-09-16

Devin adds macOS VMs, enabling iOS development without a Mac

Official

Reimagining advertising with AI

OpenAI Blog ・ 2026-09-16

OpenAI brings Sponsored Agents and CRM ties to ChatGPT Ads

Community

Salesforce AI Force, Agents as UI, The Race to Headless

Stratechery (free posts) ・ 2026-09-16

Stratechery: Salesforce drops UI as a moat in the race to headless

Press

「APIキーは.envに」はもはや通用しない AIエージェントの“内通者化”をどう防ぐ?

ITmedia AI+ ・ 2026-09-15

Beyond .env: a three-layer credential defense for AI agents

Community

There's a 100% Chance AI Agents Are Ruining the Internet

Hacker News (Front Page) ・ 2026-09-15

404 Media: AI agents are already ruining the internet

Community

Planning with Agents: Divided Worlds, Boundary Objects, and Thicker Interfaces

Lobste.rs (AI tagged) ・ 2026-09-15

GitHub Next argues agent planning needs thicker, visual interfaces

Academic (arxiv etc.) 18 ▾
Academic

Agentic Societies Need a Social Harness

arXiv cs.AI (Artificial Intelligence) ・ 2026-09-15

Academic

ScienceBuddy: Recursive-in-Recursive Self-Improvement for Interactive Scientific Agents

arXiv cs.AI (Artificial Intelligence) ・ 2026-09-15

Academic

ENCP: Episode-Normalized Conformal Prediction for Vision-and-Language Navigation

arXiv cs.AI (Artificial Intelligence) ・ 2026-09-15

Academic

Verifiable Social Reasoning for LLM Assistants

arXiv cs.AI (Artificial Intelligence) ・ 2026-09-15

Academic

Decomposition Buys Integrity, Not Yield

arXiv cs.AI (Artificial Intelligence) ・ 2026-09-15

Academic

Evaluating Verified Autonomy in Quantum Engineering

arXiv cs.AI (Artificial Intelligence) ・ 2026-09-15

Academic

Coding Agents Have Converged: Why the SWE-bench Leaderboard Can No Longer Order Its Top Entries, and What to Measure Instead

arXiv cs.AI (Artificial Intelligence) ・ 2026-09-15

Academic

FlashVector: Agent for Hierarchical Model Serving Stack Optimization

arXiv cs.AI (Artificial Intelligence) ・ 2026-09-15

Academic

Self-Emergence Agent Architecture:Behavior-Inertia HMM, Reflexive Metacognition,and Social-Contrastive Self-Modeling

arXiv cs.AI (Artificial Intelligence) ・ 2026-09-15

Academic

Intrinsic Motivation in Reinforcement Learning: A Research Agenda for Adaptive Self-Organisation

arXiv cs.AI (Artificial Intelligence) ・ 2026-09-15

Academic

Towards Detecting AI-Assisted Responses in Online Surveys

arXiv cs.CL (Computation and Language) ・ 2026-09-15

Academic

After the Party: Governing What a Viral Agent-Skill Ecosystem Left Behind

arXiv cs.AI (Artificial Intelligence) ・ 2026-09-15

Academic

Grounding SWE-Agent Decisions in Architecture-0 Design: Navigating Unknown Unknowns through Physical Mapping

arXiv cs.AI (Artificial Intelligence) ・ 2026-09-15

Academic

AI for Science with GPT-6 Astra: Thermal Design and Electrothermal Analysis of 2D CFET

arXiv cs.AI (Artificial Intelligence) ・ 2026-09-15

Academic

Symbolic Separation: Grounding Deep Agents in Knowledge Graphs for Trustworthy Operational Data Analytics

arXiv cs.AI (Artificial Intelligence) ・ 2026-09-15

Academic

Interactive Memory Learning for Long-Term Conversations

arXiv cs.AI (Artificial Intelligence) ・ 2026-09-15

Academic

ThinkFlow: Self-Evolving Probabilistic Latent Memory for Lifelong Conversational Agents

arXiv cs.CL (Computation and Language) ・ 2026-09-15

Academic

PaperDoctor: Evidence-Grounded and Actionable Feedback for Scientific Papers in Progress

arXiv cs.CL (Computation and Language) ・ 2026-09-15

← Story Archive