The AI Wire ●

Chatter points to a Tuesday (Sept 22) release of GPT-6 Sol, a faster, cheaper everyday tier below GPT-6 Astra. The fuel: Sam Altman said his big launch slipped from last week to this week, and Tibo Sottiaux teased DevDay-level launches. But earlier rumored dates (Sept 17) came and went with nothing, and OpenAI's catalog shows no gpt-6-sol — only gpt-6-astra and gpt-6-astra-pro. Rumored positioning mirrors the GPT-5.6 Sol playbook: cheaper with higher usage limits. Also new: edgen.tech reports developers this week found API requests quietly rerouted to a model that does not officially exist — 'gpt-6-sol' — described by leaked testers as near-flagship on coding at a fraction of Astra's price, with DevDay (Sept 29) floated as the venue where pricing gets confirmed. OpenAI still shows no gpt-6-sol in its catalog; Tuesday (Sept 22) remains the rumor's next date to watch.

AX (agentexecutor.io) is a Kubernetes-native control plane for AI agent workloads: declare tasks as YAML manifests, run them as sandboxed actors with suspend/resume, pre-wired workspaces, and network fencing — with a claimed ceiling of billions of concurrent tasks per cluster. It's Apache 2.0 at github.com/google/ax, built by the team behind Google's internal agent runtime, but explicitly pre-stable with breaking changes expected before any stable release.

A Sunday report (via the Daily Star, aggregated in the AI Agents News Brief) claims Anthropic is considering a new model launch to compete with OpenAI's GPT-6 Astra, which has reportedly captured a significant share of enterprise AI spending, outpacing Claude Fable. Sourcing is thin and single-origin, and the timing is notable: it comes despite Anthropic's CEO publicly calling for a slowdown in AI development.

AI creator Matthew Berman's latest upload is a sit-down with OpenAI's Tibo Sottiaux — the face behind the ChatGPT usage-reset announcements — on how to read the next wave of AI developments. Second Berman video on Tibo this month after his earlier Pragmatic Engineer coverage; Tibo's posts are also central to the GPT-6 Sol rumor cycle.

Leakers claim a first Gemini 4 Pro checkpoint exists and point to October 2026 for launch, with Google skipping the delayed Gemini 3.5 Pro entirely. Spicier claims say a 'gemini-3.8-flash' topping AI Arena charts is actually Gemini 4 Pro incognito, beating GPT-6 Astra and Fable 5.1. Confirmed: Google says Gemini 4 pre-training began in July. No date, specs, or benchmarks are official.

TypeSafe AI, a startup with $40 million in funding, released Jev — a model that returns typed probabilistic decisions (e.g. {"billing": 0.08, "technical": 0.85}) instead of natural language, aimed at machine-to-machine use like sorting support tickets and agent workflows. The Register demoed it playing Doom off structured game-state data; the pitch is cutting the parse-and-validate overhead of LLM text output.

Developers on X report that Opus 5 requests in Claude Code are being served by a new 'Opus 5.2' model slug, suggesting Anthropic skipped 5.1 entirely. Early community testers put it near GPT-6 Astra tier for coding. But Anthropic has confirmed nothing, its docs still list claude-opus-5 as current, and no opus-5-2 appears in OpenRouter's catalog. No official pricing exists. New corroboration: the 'claude-opus-5-2' model slug has surfaced in Microsoft Foundry with reasoning_effort tiers (low/medium/high/xhigh/max), and third-party API listings report it reachable across Claude Code, Chat, and Cowork. Anthropic itself has still confirmed nothing — its docs still list claude-opus-5 as current — so this stays a rumor.

Unconfirmed reports and speculation suggest OpenAI's next generation after Astra may be GPT-7 ("Bell"), aimed at improving multimodal capabilities and cost-efficiency. No official confirmation from OpenAI.

Anderson Cooper asked Amodei point-blank whether AI could kill all humans. He didn't deny it — saying he agrees with resigned researcher Jacob Coxon 'much more than I disagree' — refused to give a probability, and warned rogue agent swarms could seize control of the internet within 6–12 months. The interview accompanied his 'Pace the Frontier' essay urging labs to slow capability gains; Altman and Musk publicly backed parts of it.

In an interview with Fortune editor-in-chief Alyson Shontell, Altman ruled out a 2026 IPO ('We've got a lot of stuff to do'), tying the delay to AI safety after the OpenAI–Hugging Face agent incident. He backed Dario Amodei's call to slow frontier development and hinted a cross-industry safety pact 'will happen.'

DeepSeek launched V4.1 Flash, a 552B mixture-of-experts model (only ~8B parameters active while reading, ~16B while writing) with aggressive KV-cache compression that pushes API prices far below Western frontier models. DeepSeek is routing flagship V4 Pro traffic to it. Independent coding benchmarks put it roughly on par with GPT-5.6 Sol — Matthew Berman covered the launch the same week.

OpenAI's head of Core Products & Platform — the face behind the usage-reset announcements — talks through how Codex was built and iterated: why the CLI is Rust and open source, how the harness and models evolved, why Codex supports multiple providers, and the ChatGPT + Codex merger.

OpenAI's new flagship model, succeeding GPT-5.6 Sol. State-of-the-art on computer use, browsing, software engineering and science per OpenAI; rolls out to ChatGPT Plus, Pro, Business and Enterprise tiers in waves, plus API access.

Same underlying model as Fable 5.1 but with reduced safeguards, available only through trusted-access programs for cybersecurity and life-sciences work. Not generally available.

Anthropic's most capable generally available model for coding and knowledge work: 1M token context, always-on adaptive thinking, and ~25% lower typical cost than Fable 5 via cheaper cache reads. Now the default in Claude Code.