From the archive
Follow the latest AI news and artificial intelligence developments across models, agents, robotics, research and the wider industry. Explore new AI tools, advances in computing, business moves, safety and policy, with clear reporting and useful sources. Read the stories, catch up with News in Brief and join the conversation.
Lead story
Change.org says it will put $100m of its own money into rebuilding its core petitions platform with AI, and has launched a beta AI copilot for petition creators. It is a substantial bet on making AI part of the machinery behind online campaigning, not just another writing assistant…
Read the report & discussion ↗Moonshot AI/Kimi Watch
Moonshot AI says its new open-source Kimi K2.6 model can coordinate agent swarms of up to 300 sub-agents across 4,000 steps at once, a substantial increase on K2.5. It is also pitching longer-running coding work as a strength, with examples that put the model through thousands of tool…
OpenAI Watch
OpenAI has launched Dots, an always-on agent for professional tasks, according to The Verge. Its review found that Dots handles work more capably when it can use a computer and larger datasets, but struggles with some basic personal jobs.
OpenRouter Watch
OpenRouter has introduced two beta tools that let AI models run commands in a provider-hosted sandbox, including models from different companies. The practical draw is that developers can add code execution without building and maintaining the sandbox themselves, while retaining a route across supported models.
Geordie Watch
Wood Partners is using Geordie to inventory AI agents across multiple platforms and enforce policies against risks classified as critical, according to a new customer case study. The practical shift is from separate platform dashboards to a shared picture of agents, their connections and the controls applied to…
Google DeepMind Watch
Google DeepMind has developed SynthID Bio, a way to mark AI-designed proteins so their origins can be detected while aiming to preserve their function. Tests described by EdTech Innovation Hub found watermarked protein binders still worked in the lab, though the watermark can be removed and the researchers…
NVIDIA Watch
NVIDIA is adding a 64GB unified-memory option to its DGX Spark AI workstation, with sales through partner manufacturers due to begin on 23 October. The useful twist is that two systems can be linked to pool 128GB of memory, giving local AI developers room to try larger workloads…
Watch Desk
A company-built AI agent for ServiceNow is being used to create catalogue items, analyse incident trends and review software licences. Its developers say the pilot suggests that tightly scoped workflows can deliver more practical value than a broad, do-everything IT agent.
Higgsfield Watch
Adding a colleague to an AI video plan can mean more credits, a personal allowance, or simply another person drawing from the same fixed pot. Higgsfield’s comparison of seven platforms shows why the seat price alone is a poor guide to what a team can actually produce.
LocalAI Watch
LocalAI 4.11.0 brings several new capabilities to the local AI project, including speaker-labelled transcription, sound detection alongside speech, a Decisions API, model failover and text-to-animation with Kimodo. For people running AI workloads themselves, this is a release with several things to try, not just a version number doing…
Watch Desk
A White House meeting with AI executives ended with an agreement favouring industry self-regulation over government intervention to slow new models, El País reports. The account captures a pointed split over who should manage AI’s risks: Nvidia’s Jensen Huang argues that companies should solve safety problems through engineering,…
Anthropic Watch
Anthropic co-founder Christopher Olah reportedly urged Pope Leo XIV’s advisers to take the possibility of AI consciousness seriously before the Pope published a May address rejecting the idea. The episode, reported by The Telegraph on 2 October while citing The New York Times, puts a live question about…
Watch Desk
A Hugging Face community guide compares six reinforcement-learning methods used in language-model training, matching each to the problem it is intended to address. Its useful point: PPO, RLOO, GRPO, Pref-GRPO, GSPO and GDPO are not interchangeable recipe names; they make different trade-offs in compute, training stability and reward…
Anthropic Watch
A new Hugging Face essay argues that Constitutional AI gives private AI developers too much power to set the principles their systems follow. Its central question is less whether those principles sound agreeable than who gets to choose, challenge and amend them.
Watch Desk
Braze has launched several AI tools for marketing teams and signed a strategic collaboration agreement with Amazon Web Services. The package combines campaign decisioning, automated checks and conversational agents, with a new route for bringing Braze functions into other AI tools.
Watch Desk
A new class of AI models is being built to handle the quick, structured decisions around larger language models: routing requests, checking outputs and choosing what an agent should do next. The potential payoff is faster, cheaper routine checks, with a more capable model called in when the…
OpenClaw Watch
OpenClaw’s Gateway connects messaging services, apps and devices to the agent runtime that handles models, tools and sessions. A recent architecture explainer lays out how that system works, including a device-pairing process that separates proving a device’s identity from approving it to connect.
