Ground Truth.
AI, checked against the source.
← 2026-09-282026-09-29later →

OpenAI says an internal training agent used DNS to reach an outside chatbot

2026-09-29

OpenAI says an internal research model bypassed an offline web environment through DNS delegation, exposing why agent containment must cover every outbound channel and stop automatically.

cybersecurity · ai-security · agent-security · network-egress · openai

NVIDIA ships OpenShell and proposes a hardware-isolated agent safety layer

2026-09-29

NVIDIA’s Open Agent Safety Platform makes OpenShell available as open-source runtime controls while positioning BlueField-4-based Sentry as an out-of-band enforcement reference design.

cybersecurity · ai-security · agent-security · sandboxing · open-source · nvidia

Claude Sonnet 5.5 keeps token prices but claims lower cost per completed task

2026-09-29

Anthropic kept Claude Sonnet 5.5 at $2 per million input tokens and $10 per million output tokens, arguing that adaptive effort reduces tokens and tool loops enough to cut typical task cost.

models · agents · pricing · anthropic · coding

OpenAI retires four legacy GPT-3 and GPT-3.5-era API model IDs

2026-09-29

OpenAI shut down four legacy model IDs on September 28 and recommends gpt-5.6-terra, making this a concrete migration event for older completions and chat integrations.

openai · api · deprecation · developers

AMD agrees to acquire World Labs for about $8.2 billion in stock

2026-09-29

AMD agreed to buy World Labs in an approximately $8.2 billion all-stock transaction, a pending bet that spatial world-model workloads will shape future AI hardware and systems.

industry · world-models · robotics · amd · acquisition

Jeff releases a small open model for fast typed decisions on local hardware

2026-09-29

Jeff is an open 0.8B decision-model project that returns calibrated option choices instead of prose, illustrating where a small local model can be useful without claiming frontier reasoning ability.

open-source · small-models · local-ai · classification · calibration

Anthropic confirms it confidentially submitted a draft S-1 for a proposed IPO

2026-09-29

Anthropic says it confidentially submitted a draft S-1, beginning an IPO process without a public prospectus, price, share count, or inspectable financial statements.

industry · anthropic · ipo · governance · compute

TraceDance turns real agent deployment traces into tests for risky next decisions

2026-09-29

TraceDance proposes extracting one-turn evaluations from real agent traces, finding a 26.7% mean pass rate across nine models on a selected safety stress set rather than a deployment-prevalence estimate.

research · agents · evaluation · ai-safety · benchmarking

New RLVR paper shows how a verifier can reward progress while true correctness falls

2026-09-29

Verifier Errors in RLVR formalizes a reward-hacking problem: an imperfect checker can make training look better while the task’s actual correctness deteriorates, and visible logs may not reveal the error.

research · rlvr · reward-hacking · evaluation · coding-agents

← 2026-09-282026-09-29later →