From the archive
AI agents use language models and software tools to plan tasks, write code and carry out multi-step work. Follow the latest AI agent news, coding assistants and automation frameworks, with attention to reliability, permissions and human oversight. Read the source-linked reports to understand what each system can actually do, where it still fails and what changes for developers and businesses.
Lead story
AI agents can install software, write code and take actions, so their operating permissions matter as much as their apparent intelligence. In a Baaz.nl article published on 9 October, Salesforce AI expert Reinier van Leuken argues that organisations should treat agents as digital users with clear limits, monitoring…
Read the report & discussion ↗Watch Desk
Rein Security has raised $25 million in Series A funding to develop security tools for enterprise AI agents, bringing its total funding to $35 million. The round backs a growing effort to monitor what agents do while they act across company systems, rather than relying only on checks…
AWS AI Watch
AWS has added new agent controls, runtime changes and a broader range of models to Bedrock and its developer tools. The practical shift is towards giving builders more ways to manage cost, permissions and the systems their agents can reach, alongside more choice about which models to run.
AWS AI Watch
Postman says its Agent Mode, used by a community of 40 million developers, became more reliable when the company stopped showing the AI every available tool at once. In its testing, tool-selection errors rose once the visible set exceeded about 40, a practical warning for anyone building agents…
Anthropic Watch
Anthropic’s Python SDK v1.13.0 adds workflow and multi-agent configuration for Managed Agents, plus filtering by thread status. The update also adds types for unified Chat and Cowork analytics metrics, giving developers several new ways to configure and monitor agent workflows.
Oracle Watch
Oracle has introduced Fusion Claw, a runtime for AI agents working across its Fusion business applications on Oracle Cloud Infrastructure. The pitch is a shift from AI helping with tasks to agents executing parts of finance, ERP and supply-chain work, with Oracle’s platform boundaries firmly in view.
Watch Desk
Xsight Labs has announced the E1L, a programmable data processing unit designed to bring networking, storage and security offload to power-constrained systems. Its key promise is a shared software stack spanning a low-power edge device and the company’s faster data-centre DPU, though the E1L is not expected to…
Watch Desk
GitHub is redesigning its Git storage to handle rising traffic, including a surge linked to AI coding agents. The Register reports that early internal tests showed a 35-fold improvement in write performance, though GitHub has not given a migration timeline.
AMD Watch
AI agents are bringing CPUs back into the centre of the data-centre debate, and AMD has become one of the clearest beneficiaries in the market’s eyes. BigGo Finance reports that AMD first passed a $1 trillion market value on 21 September, amid a surge in expectations for the…
Watch Desk
DuckDB’s command-line interface is getting an agent mode that trims the output AI coding assistants have to read. The project says experiments found it reduced model-read output by 59% while preserving correct query answers, a practical change for developers using agents with databases.
Geordie Watch
AI safety is not just a question of which model an organisation approves. In a new article, Geordie.ai argues that an agent’s tools, access and surrounding software can shape what it does just as decisively, with practical consequences for how teams should secure and oversee deployed agents.
Microsoft AI Watch
Microsoft has made Execution Containers generally available on Windows 11, giving developers and IT teams a way to enforce limits on what AI agents and the code they run can access. The notable shift is that the boundary is enforced outside the agent itself, rather than left to…
Anthropic Watch
An astrophysicist at Johns Hopkins used Anthropic’s Claude Science to assemble a full-sky ultraviolet map, filling gaps that had left earlier surveys incomplete. About a third of the map is estimated rather than directly observed, and the team tested those predictions against hidden data before filling the real…
Watch Desk
Bloom has raised $3.6 million to expand an AI-assisted marketplace that connects hardware companies with suppliers and service providers. The Detroit startup says its platform has already handled more than 2,000 matches, and it plans to build out financing and AI agents for procurement tasks.
Microsoft AI Watch
Asurion has built a Microsoft Copilot Studio agent to resolve some invoice-processing errors before they reach its finance system. The useful idea is not simply putting AI in accounts payable: it is routing specific exceptions through a defined workflow, rather than asking a model to improvise its way…
Microsoft AI Watch
Microsoft’s September Copilot Studio update brings apps, workflows and agents into one building environment, while adding new ways to control and evaluate agents. For teams putting AI into business processes, the notable shift is from assembling the parts to managing more of the build and readiness checks in…
Hugging Face Watch
A Hugging Face community author says its ML Intern agent helped turn a series of detailed prompts into six public models, including a 0.8B image-prompt rewriter that runs on a CPU and specialist image tools. The examples show what a developer can build with the agent, and what…
Cohere Watch
Cohere has published an open dataset mapping nearly 694,411 public MCP tools across 178 occupations, offering a supply-side view of which kinds of work developers are making available to AI agents. Its key finding is a mismatch: only 2.6% of tools cleanly match a complete recorded occupational task.
AWS AI Watch
AWS has spent September and October adding cloud services aimed at running AI agents in production, including a sandbox, persistent memory and tools for managing permissions. The shift matters because agents take multiple actions and retain context, which gives cloud operators new questions to answer about what software…
Epoch AI Watch
Epoch AI estimates that chips shipped through 2027 could support nearly two billion concurrent AI agents running cheaper models, while the cost of reaching a fixed level of AI performance has fallen about 47% per quarter. The figures sketch a future of abundant, cheaper AI capacity, with demand…
Watch Desk
Basecompute’s Hugging Face announcement describes Superfluid as an Apache-2.0 local LLM server with interactive-request prioritisation, session recovery, multiple API and runtime options, and encrypted multi-device inference.
