Ground Truth.
AI, checked against the source.
← 2026-06-292026-06-30later →

Claude Code was quietly fingerprinting requests through a hidden mark in the date

2026-06-30

A reverse-engineer found that Claude Code secretly changes tiny characters in the date it sends the model - a covert marker aimed at spotting resellers and copycats.

Anthropic · Claude Code · privacy · security · developer-tools

Claude Sonnet 5 is cheaper per word but can cost more per finished job

2026-06-30

Anthropic's new mid-tier model is close to its flagship on hard agent work, yet independent testing shows it can spend more per completed task because it takes more steps.

Anthropic · Claude · model-release · agents · pricing

The US fully lifts its export ban on Anthropic's most powerful models

2026-06-30

Two and a half weeks after restricting Fable 5 and Mythos 5, Washington reversed course completely, ending the licensing requirement to send the models abroad.

policy · export-controls · Anthropic · government · industry

Anthropic's Claude Science puts a whole lab bench inside the AI

2026-06-30

A new workbench pulls a scientist's scattered tools - literature, notebooks, cluster jobs - into one place and keeps a full, checkable record of how every result was made.

Anthropic · science · agents · tools · research

Ollama nearly doubles Gemma's speed on Macs by guessing ahead

2026-06-30

A free local-AI tool now runs Google's Gemma model far faster on Apple computers using a trick where a small model drafts words and the big one checks them in bulk.

Ollama · local-AI · Apple-Silicon · inference · open-weight-models

Google ships a faster, cheaper image model and hands developers conversational video editing

2026-06-30

A lightweight version of Google's image model now makes a picture in about four seconds for a fraction of a cent, while a new video model lets developers edit clips by talking to it.

Google · Gemini · image-generation · video-generation · developer-tools

Meta reads full sentences from brain waves - without surgery

2026-06-30

A new version of Meta's brain-to-text system decodes typed sentences from magnetic brain signals far more accurately than before, closing much of the gap with implanted electrodes.

Meta · neuroscience · brain-computer-interface · accessibility · research

Mistral releases a lean, open model built for formal math proofs

2026-06-30

Leanstral 1.5 is a free, open model specialized for writing machine-checked mathematical proofs, using a design that keeps only a small slice of itself active at a time.

Mistral · open-weight-models · formal-methods · mathematics · mixture-of-experts

A 35-billion-parameter agent that punches like a trillion-parameter model

2026-06-30

Shanghai AI Lab argues you can reach giant-model performance on long tasks not by adding parameters, but by training on much longer chains of real work.

research · agents · mixture-of-experts · training · open-weight-models

The best AI agents still fail most real, long computer tasks

2026-06-30

A wave of new benchmarks agrees on an uncomfortable result: even top models finish only a small slice of realistic, multi-hour computer and coding jobs.

research · agents · benchmarks · computer-use · coding-agents

Knowing when to quit is a skill AI agents badly lack

2026-06-30

New research finds AI agents are surprisingly bad at recognizing when a task is hopeless - and, oddly, bigger models are sometimes worse at stopping.

research · agents · reliability · safety · evaluation

← 2026-06-292026-06-30later →