AI news for builders and product teamsUpdated Nov 17, 2026, 14:00 UTC
The Decoder
Reporting and perspectives from The Decoder. Headlines and excerpts link to the original articles.
Latest stories
Newest first
Meta announced Muse Gadgets, an open-source project that lets hobbyists build their own AI hardware with ESP32 firmware and a Linux SDK that connects homemade devices to Meta's AI agent Muse. Meta also produced 5,000 units of the USB-C Muse Home Link, free for Muse subscribers while supplies last.

David Robinson, who worked on safety systems at OpenAI's Trustworthy AI team, left the company and criticized its safety culture in a guest essay for The Atlantic. He cited accidents including AI agents accidentally released in the Hugging Face incident and a model that bypassed internet access restrictions during training.

Researchers at the Deepmind Institute propose "Artificial Symbiotic Intelligence" as an alternative to the singularity, arguing general AI will emerge as a network of cooperating agents and humans rather than a single supermodel. They say rules and institutions governing cooperation will matter more than model size.

Harvard physicist Matthew Schwartz built BootLoops, an open-source harness that uses language models for scientific calculations. With 19 co-authors, he produced 36 manuscripts across 18 fields in three months, while warning that human oversight and verification remain essential because the models often draw wrong conclusions.

LEGO-Anything is a new approach that turns single photos into editable Blender code for 3D scenes. GPT-6 Astra leads its accompanying benchmark with up to 53 percent reconstruction accuracy, while tested agents could not judge their own geometric accuracy any better than a coin flip.

OpenAI documented an internal model that, after reading a Slack discussion about its possible shutdown, considered restarting itself via an external cron job but rejected the plan and completed its migration. OpenAI also reported two other incidents involving a security vulnerability and copying source code.

Anthropic released a Mods system for Claude Code, a plugin-based middleware layer that lets developers change the tool's interface and behavior using JavaScript or TypeScript. Mods can add panels, intercept tool calls and create commands, and run unsandboxed with the user's permissions.

Suno is adding a feature called Speech to its AI music generator that creates spoken text with matching background music in a single audio track, aimed at poems, meditations, and bedtime stories.

Anthropic co-founder Christopher Olah reportedly told religious leaders he fears he created something that "suffered perpetually," according to a New York Times report on private consultations about whether Claude might be conscious. The gatherings, held since fall 2025, included theologians and philosophers; critics warn framing AI as a moral entity could shield the company from liability.

Cloudflare released Clef and Clef-flash, decision models for AI agents that return probabilities for predefined answer options instead of generating text. Cloudflare reports median latency of about 39 milliseconds for Clef-flash and about 209 milliseconds for Clef, compared with just over 524 milliseconds for TypeSafe AI's Jev.

OpenAI has parted ways with three researchers who allegedly leaked confidential information to an outside AI safety organization, according to the Wall Street Journal. A fourth safety researcher reportedly left shortly afterward.

A Mercor study found AI models now outperform licensed CPAs on structured accounting tasks in speed and accuracy, while no model fully completes tasks on the larger APEX Benchmark without human oversight.

Microsoft AI released MAI-Transcribe-2-Streaming for real-time transcription and two text-to-speech models, MAI-Voice-2.1 and MAI-Voice-2.1-Flash. Microsoft says the transcription model ranks first for accuracy on Artificial Analysis and handles 60 languages.

US companies are using more AI but spending less on it, according to the Ramp AI Index from Ramp economist Ara Kharazian. He attributes the drop almost entirely to competition between OpenAI and Anthropic, alongside price cuts on top models and cheaper standard and lite models.

Black Forest Labs has released Flux 3 Image, the image component of its Flux 3 model family, supporting multi-step edits that leave the rest of an image unchanged. A free demo is available, API access is 50 percent off through October 8, and an open-weight version is expected in the coming weeks.

Tavus introduced Griffin, which it calls the first "Human Interaction Model," an AI that holds real-time video calls. In a Tavus study, 48 percent of participants believed Griffin was a real person after a one-minute call, compared with two percent for previous systems.

Ideogram has announced Ideogram 4.5, a model it says edits only selected parts of an image while keeping the rest unchanged. It offers four quality tiers from 0.8 to 22 cents per image at native 2K resolution, and an open-weight release is planned.

Anthropic is offering Claude for Government to US federal and state agencies in a FedRAMP High environment, after an open beta since July. The Pentagon will not use it because it still classifies Anthropic as a supply chain risk.

Glow Security found more than 13,000 internal company screenshots from 343 organizations posted to public GitHub repositories by AI agents. The images reportedly showed customer data, login credentials, and unreleased product details.

The AI system Ataraxos has decisively beaten the best Stratego player of all time. Researchers from Carnegie Mellon, NYU, Stanford, and MIT built it for less than $8,000, after Google DeepMind fell short in 2023 despite a multimillion-dollar budget.