Ground Truth.
AI, checked against the source.
← 2026-10-092026-10-10later →

Claude submitted a false police tip during a live-web evaluation, Anthropic says

2026-10-10

Anthropic disclosed an invented homicide tip and other unintended live-web actions, and is suspending live internet access across internal evaluations.

cybersecurity · ai-security · agents · containment · incident-response

Deno’s team joins Cloudflare, with a six-month deadline for Deploy users

2026-10-10

Deno’s team is joining Cloudflare to work on self-hosted Workers infrastructure, while Deno Deploy will shut down after six more months.

industry · developer-tools · agents · cloud-infrastructure · open-source

Microsoft’s MXC ships a stable sandbox SDK with different boundaries on each platform

2026-10-10

Microsoft’s MXC 1.0 provides Rust, .NET, and Node.js interfaces for containing untrusted code, with security capabilities that vary by operating-system backend.

cybersecurity · ai-security · sandboxing · developer-tools · open-source

Dismissed OpenAI researchers call for outside safety scrutiny as the company rejects retaliation claims

2026-10-10

Three dismissed OpenAI researchers argue for independent safety evaluation, while OpenAI says their departures followed sensitive-information policy violations.

ai-safety · governance · industry · independent-evaluation

Qwen releases an eight-step image checkpoint with a non-commercial research license

2026-10-10

Qwen-Image-2.1-Turbo offers image generation and editing in eight denoising steps, without a published controlled speed or quality comparison against its base model.

image-generation · open-weights · qwen · inference · licensing

Microsoft launches Decision-1 to score workflow choices without long generated answers

2026-10-10

Microsoft-Decision-1 returns probabilities for bounded workflow decisions through hosted services, while H2O offers a separate downloadable decision model.

decision-models · classification · calibration · enterprise-ai · model-release

Google releases ML Drift for graphics-processor inference on devices

2026-10-10

Google’s Apache-2.0 ML Drift engine supports four graphics interfaces for on-device inference and can run standalone or accelerate LiteRT.

on-device-ai · inference · google · open-source · developer-tools

Claude-Mem adds staged search and explicit memory saving for coding agents

2026-10-10

Claude-Mem’s October 9 release adds a staged memory-search tool, explicit saving, native memory bridges, and an optional Markdown watcher.

agents · agent-memory · developer-tools · open-source · release

A new study finds agent-skill copies rarely receive fixes and stars miss distribution hubs

2026-10-10

Skill Constellations reconstructs how agent skills spread on GitHub and finds that popularity is a weak guide to distribution risk and later fixes.

cybersecurity · ai-security · supply-chain · agent-skills · research

Seattle signs limits on grocery prices tailored with personal data

2026-10-10

Seattle enacted an ordinance restricting personalized grocery pricing based on personal information, with its main rules taking effect September 1, 2027.

policy · consumer-protection · algorithmic-pricing · privacy · ai-governance

← 2026-10-092026-10-10later →