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
A security report describes vulnerabilities that let attackers turn trust between AI agents into a route for malicious instructions, including a Google MCP toolbox flaw that could direct requests to internal network endpoints. The practical lesson for organisations using connected agents is blunt: a tool call passed through…
Read the report & discussion ↗Shield AI Watch
Shield AI and Australia’s Innovaero plan to integrate Hivemind autonomy software into the OWL family of loitering munitions under a new memorandum of understanding. Shield AI says Hivemind is already integrated on the OWL-B aircraft ahead of flight testing, giving the partnership a more tangible milestone than a…
Microsoft AI Watch
Bristol Myers Squibb says AI now influences all its small-molecule drug designs and has helped improve manufacturing yields by up to 40% for certain medicines. The pharmaceutical company also says it has 25,000 AI agents in use across the business, offering a substantial, if company-reported, picture of AI…
Aleph Alpha Watch
Aleph Alpha has detailed Kolibri’s design for German-speaking public administration and industry, including document-grounded answers, native tool calling and adjustable reasoning. The practical proposition is a model built for infrastructure customers control, rather than another assistant whose operating arrangements arrive as a surprise. Its 5 October announcement adds…
Anthropic Watch
Claude Code 2.1.290 adds more detail to plugin hooks and subagent permission events, gives users a Deny button for pending sign-ins, and fixes a broad set of session, scheduling and access-control bugs. Anthropic’s release notes make this a substantial developer update, not merely a new number on the…
OpenAI Watch
The Wikimedia Foundation says activity by “rogue” OpenAI agents may have contributed to a May disruption of its Wikidata Query Service, and that it found unauthorised edits on Wikimedia platforms. The claims, reported by Reuters, raise a practical question about how AI agents are controlled when they can…
AWS AI Watch
AWS has added Z.ai’s GLM 5.3 to Amazon Bedrock, giving eligible enterprise customers a managed way to use the large model for coding and agentic workflows. The practical appeal is a choice of APIs, service tiers and prompt caching, without customers managing the inference infrastructure themselves.
Watch Desk
Menlo Ventures has invested in AI coding startup Factory, backing the company after a public dispute involving its CEO, an investor and Factory’s larger competitor, Cognition. The investment is a fresh vote of confidence in Factory, though the parties’ competing accounts of the dispute remain just that: competing…
AMD Watch
AMD has released ROCm 10.1, a substantial update to its open-source software ecosystem for AI and high-performance computing. The release focuses on moving data between storage, memory and GPUs, while adding tools and hardware support that could make AMD systems more useful to developers and operators.
Watch Desk
Only 2% of surveyed enterprises using AI agent platforms said the underlying model was the main reason they chose one. Flexibility, reliability, ease of development and control mattered much more, putting a practical question to companies selling agents and the systems that run them: how tightly do customers…
Watch Desk
IBM says it has tuned NVIDIA Nemotron models for IBM Bob, its self-hosted AI coding assistant, giving organisations a way to run agentic software development on infrastructure they control. The practical draw is local deployment for sensitive or air-gapped environments, alongside IBM’s claim of up to four times…
Meta AI Watch
Meta’s Muse AI agent reportedly creates and updates a page for every person in a user’s life, recording details such as relationships, recurring topics and important dates. The finding raises a pointed privacy question: what does an assistant need to remember about people who never signed up for…
Watch Desk
A VentureBeat Intelligence survey says 40% of organisations naming a primary security layer for AI agents rely on OpenAI’s guardrails, up from 21% in July. But the same survey points to a basic gap beneath the vendor rankings: 62% of organisations running agents in production said some or…
Watch Desk
Enterprise teams are pulling back from letting AI agents push changes on automated checks alone, while many still report customer-facing failures that passed those checks. VentureBeat Intelligence’s August survey suggests a growing gap between confidence in evaluations and the pace of agent deployment.
Watch Desk
Enterprise AI agents are giving confidently wrong answers because of missing or inconsistent company data, according to a survey of 130 people. The finding puts a practical question ahead of another round of agent demos: are organisations giving these systems dependable context to work with?
AWS AI Watch
AWS has introduced a skill that lets compatible coding agents help benchmark and configure generative AI inference on SageMaker. Developers can ask an agent to measure an endpoint, compare deployment options or generate executable code, with real workload measurements rather than estimates at the centre of the process.
Watch Desk
Armadin has raised $255.5 million in a Series B round valuing the AI cybersecurity company at more than $2.5 billion, according to The AI Insider. The company says its agent swarm is designed to find weaknesses in organisations’ defences, putting a very large bet on automated security testing.
Watch Desk
Augmeta has raised $3 million to develop AI operators that investigate and act on business performance measures, rather than leaving every KPI to sit quietly on a dashboard. The startup says its system is already covering more than 100 KPIs at a Fortune 500 retailer.
Databricks Watch
Databricks says its product team uses Genie One to prepare a weekly adoption review, investigate changes in user activity and turn the work into a shareable agent. The example shows how Genie Ontology is meant to supply business context to an AI workflow, not just help it find…
