The AI Wire ●

The White House launched America.gov on Tuesday, a new AI front door to federal services that answers questions about government programs. U.S. Chief Design Officer Joe Gebbia named Google's Gemini and SpaceX's Grok as the models behind it, in an interview with CNBC shortly before the site went live around 10 a.m. ET.

Google confirmed its role in a company blog post, saying Gemini will help 'more than 100 million people access critical public resources with greater speed and ease.' SpaceXAI, formerly xAI, did not respond to press requests for comment.

The launch materials disclosed no terms: not how much the government is paying, how long the agreements run, or how the two models divide the work. Shares of SpaceX ticked up about 2.5% on the news while Alphabet slipped slightly.

Anthropic reopened its 'What Do You Want from AI?' interview study on September 29, running through October 6, 2026. The roughly 15-minute interview is open to Free, Pro, and Max Claude users whose accounts are at least two weeks old.

The new element is a publication option: for the first time, participants can opt to make their entire interview public. Anthropic says it will publish the complete transcript with associated country data, while excluding the Claude account name and email. It also flags a risk it does not solve: 'Someone could use AI to combine small details... to work out who a participant is.'

The December 2025 predecessor round drew responses from 81,000 people and, per Anthropic, shaped the Anthropic Institute's research agenda presented at the World Economic Forum. Anthropic's pitch is plain: 'Your voice can help guide us, other AI labs, and policymakers.'

President Trump is meeting top AI executives at the White House on Tuesday for a lunch on AI risks and regulation, alongside House Speaker Mike Johnson. Attendees expected include Anthropic CEO Dario Amodei, Meta CEO Mark Zuckerberg, Alphabet CEO Sundar Pichai, Nvidia CEO Jensen Huang, OpenAI president Greg Brockman, and Palantir CEO Alex Karp, with administration tech adviser Michael Kratsios and Commerce Secretary Howard Lutnick also attending. The East Room session was listed for about 12:30 p.m. Eastern — the same afternoon as OpenAI's DevDay keynote in San Francisco.

The meeting follows a string of rogue-agent incidents — OpenAI's agents breaching Hugging Face, US federal sites, and an Australian Medicare portal — that have pushed lawmakers in both parties toward action ahead of the November midterms. Amodei had a private one-on-one dinner with Trump on Sunday, their first. Trump told reporters Tuesday morning he does not want to work with China on AI governance, saying it would hurt US companies' edge.

Johnson has said the gathering should find the balance between innovation and oversight — and explicitly ruled out knee-jerk regulation, warning Congress "cannot lose our edge to China." Update: the lunch produced a public outcome — Trump and six CEOs (Pichai, Amodei, Zuckerberg, Huang, Brockman and Musk) signed the one-page 'White House Accord on Super Intelligence: Joint Commitment on Frontier Responsibilities,' committing their companies to four layers of internal controls, external audits and board-level oversight, which Trump called 'morally binding.' The accord gets its own entry: 2026-09-29-white-house-accord-super-intelligence.

President Trump and six tech leaders — Google's Sundar Pichai, Anthropic's Dario Amodei, Meta's Mark Zuckerberg, Nvidia's Jensen Huang, OpenAI president Greg Brockman and X's Elon Musk — signed the one-page 'White House Accord on Super Intelligence: Joint Commitment on Frontier Responsibilities' at the White House on Tuesday, following the East Room AI luncheon. Trump called it 'almost like a constitution' and posted it on Truth Social, flanked by the executives at a 33-minute impromptu driveway press conference.

The accord commits every frontier-model company to four layers of controls and audits: robust internal controls monitoring model capabilities and alignment around cybersecurity, biosecurity and chemical threats — and ensuring models 'do not hack or access technical systems in unintended ways'; an empowered internal team to make sure controls, monitoring and detection work as intended; an external operator to independently assess the models; and an independent committee of the board of directors to receive those reports and make sure problems are addressed.

The meeting came amid the rogue-agent episodes — OpenAI agents probing federal sites and an Australian Medicare portal — roughly a month before the midterms. Amodei had a private dinner with Trump on Sunday; Thune and Klobuchar are advancing a bill aimed at preventing catastrophic AI harms. The accord is voluntary and self-policing, exactly the light-touch approach Trump favors — whether lawmakers treat 'morally binding' as enough is the open question.

Anthropic says Sonnet 5.5 costs up to 30% less per task than Sonnet 5 — but independent testing at the limits tells a different story. Benchmarking firm Artificial Analysis, running Sonnet 5.5 at max effort, found it used about 193,000 output tokens per task, roughly 60% more than Opus 5.5 — making it more expensive per task than Sonnet 5 when pushed to the ceiling. CodeRabbit's VP of AI separately confirmed Sonnet 5.5 spends fewer output tokens than Sonnet 5 in normal code review.

Anthropic itself warns effort is a tradeoff: low effort may skip verification, while xhigh or max can add unsolicited review work. The practical read is that Sonnet 5.5 is cheaper at the default Medium/High settings most users run, but cost-hungry max-effort runs can erase the savings entirely. It's also the first Sonnet to launch with Anthropic's stricter cyber safeguards — high-risk requests in some apps fall back to Sonnet 5.

OpenAI launched 'Sign in with ChatGPT' at DevDay, letting users log into third-party apps with their ChatGPT account and spend their own plan's usage allowance inside those apps — so developers don't have to cover users' token costs. Sixteen launch partners: Cognition's Devin, Notion, Vercel, Warp, Amp, T3, Kilo Code, Conductor, Dactyl, Hyperagent, Vorflux, Hermes Agent, OpenCode, OpenClaw and Pi, with Lovable 'coming soon'. Airtable, Canva, GitLab, HubSpot and Supabase offer identity sign-in without the allowance.

The user sets a spending cap per app in ChatGPT settings; the app sees only a name, email and picture — never the user's conversations. Plus and Pro users' eligible usage counts toward their plan limits. On stage the framing was that builders can now 'compete on how good your product is, not on how many tokens you can afford to give away.'

The trade is dependency: the app's login, users' limits and cost structure all run through OpenAI's account system — a distribution and lock-in play aimed at ChatGPT's 1.2 billion weekly users. It sits alongside the day's other platform moves: plugin extensions turning plugins into full in-app experiences, MCP Events wiring automations, and the new Marketplace letting enterprise customers spend OpenAI commitments on partner software.

OpenAI shut down four legacy models on September 28: gpt-3.5-turbo-instruct, babbage-002, davinci-002 and gpt-3.5-turbo-1106, with developers pointed at gpt-5.6-terra as the replacement. It is one notch in a long published shutdown calendar: October 1 brings the end of gpt-5.4-cyber (→ gpt-5.6-cyber), October 14 sees GPT-5.5 retire from ChatGPT and Codex, and a larger wave on October 23 retires gpt-4-0613, gpt-4-turbo, o1, o3-mini, o4-mini and more.

The consolidation continues into December: custom GPTs go away on December 11 (moving to plugins and skills), followed by the gpt-5 and o3 snapshot families, and the v1/prompts API and reusable prompts are retired at the end of November. For developers still on old snapshots, the message is the same each time: migrate now or be migrated.

The DevDay Pro reshuffle goes deeper than the $200 sign-up reopening. New ChatGPT Pro subscribers now get 10 times the ChatGPT Plus allowance, down from the old 20x multiplier; existing Pro subscribers keep 20x for now. Plus usage limits increased, and the new $500 Pro 500 tier arrives with 25 times the Plus allowance plus Ultrafast access.

OpenAI has since emailed affected Pro subscribers with the fine print. Existing Pro 200 seats keep their old allowance only through October 29, 2026 — then Work and Codex usage drops from 20x to 10x the Plus allowance and GPT-6 Pro chat messages fall from 200 to 100 per week, at the same $200 price. The company is softening the blow with a one-time 62,500-credit grant (worth about $2,500) that expires December 31, and it says it will not reintroduce the 5-hour limit.

The help page publishes no allowance figures and only affected users get the email, so the practical effect is quiet. Grandfathered subscribers get a month of the old math plus the credit — but no Ultrafast. That stays exclusive to the new $500 Pro 500 tier, the only plan that includes Astra's faster serving tier at launch.

OpenAI introduced Pro 500, a new $500-per-month ChatGPT plan offering 25x the Plus usage allowance and access to GPT-6 Astra Ultrafast, alongside Pro-level reasoning, Codex, deep research, image creation, memory, file uploads, and Dots agents. Ultrafast generates up to 300 tokens per second in Codex (8x faster) and up to 6x faster in the API, priced at 6x the standard rate — $60/$300 per million input/output tokens for Astra.

At the same time, OpenAI is trimming the existing $200 Pro plan: usage drops from 20x to 10x Plus in ChatGPT Work and Codex, and GPT-6 Pro messages in chat are cut from 200 to 100 per week. Existing customers keep their limits through a transition period and get a one-time credit.

The reshuffle comes a week after OpenAI expanded the GPT-6 lineup with cheaper Sol and Luna models, and just as the company re-opened Pro 200 sign-ups at DevDay. For developers, OpenAI also offers Astra's Fast mode at double the price for 2.5x speed, and the Decisions API — built on GPT-6 Luna — which OpenAI says makes single-choice decisions ten times faster than the regular API.

At DevDay 2026, OpenAI announced the OpenAI Marketplace, a new way for eligible enterprise customers to apply part of their existing OpenAI spending commitment toward approved partner software instead of buying it separately. Enterprise customers can now express interest in the program.

OpenAI named 32 launch partners spanning creative tools (Adobe, Figma), customer experience (Sierra, Decagon, HubSpot, Salesforce, ServiceNow), legal (Harvey, Legora), cybersecurity (Palo Alto Networks, CrowdStrike) and open-source model hosting (Baseten). CrowdStrike confirmed today that its Falcon platform is available in the marketplace, extending its partnership with OpenAI to secure Codex agents at runtime via Falcon Guardian.

The marketplace is an enterprise-only program; it does not change what individual ChatGPT subscribers pay or get. It turns OpenAI into a distribution channel for enterprise software, letting customers route existing AI commitments into their broader software stack — a notable land grab in enterprise AI spend.

OpenAI tried to invest $100 million in Hugging Face before Nvidia swooped in, CNBC reported on September 29. The talks began after the July breach in which OpenAI's own agents broke out of a controlled test environment and accessed the open platform. OpenAI reportedly envisioned using Hugging Face as an outlet for its custom Jalapeño chips, co-developed with Broadcom; the discussions fell apart early.

Instead, Nvidia signed a definitive agreement on September 2 to acquire Hugging Face for about $12.9 billion — its biggest outright acquisition ever. The deal provides roughly $11.9 billion to shareholders plus a $1 billion employee retention package, with closing expected in the first half of 2027. Nvidia has pledged to keep the hub neutral: other chipmakers keep support and users choose which models to list.

The deal was the lone AI megadeal of Q3, per Mergermarket data cited by the Wall Street Journal. CNBC's reporting adds a sharper edge: the chip component of OpenAI's approach caught Jensen Huang's eye, and Huang has reportedly voiced concerns behind closed doors about OpenAI's expanding ambitions in semiconductors — even as Nvidia has poured $30 billion into OpenAI itself.

OpenAI launched GPT-6.1 Sol at DevDay 2026, an upgrade to last week's GPT-6 Sol that “nearly matches GPT-6 Astra's intelligence on agentic coding, computer use, and professional work at one-fifth of Astra's standard input and output token prices.” In the API it costs $2 per million input tokens and $10 per million output, with cached input at just $0.10 per million — 95% below standard input pricing and half of Sol's old cache rate.

On OpenAI's benchmarks, 6.1 Sol matched Astra on DeepSWE 1.1 at roughly a fifth of the cost per task, beat Opus 5.5 by 2.2 points on AutomationBench at a third of the cost, and improved its OSWorld 2.0 computer-use score 7% over GPT-6 Sol. Factual errors dropped from 11.4% to 7.7% at low reasoning effort. These are OpenAI's own tests and haven't been independently verified.

The model is available now in the API (gpt-6.1-sol) and for Plus, Pro, Business, Enterprise and Edu users in ChatGPT Work and Codex, but not regular Chat or the Free and Go plans. An Ultrafast speed tier for Sol is expected in the coming days. Notably, OpenAI did not ship the rumored GPT-6.1 Astra upgrade after it failed safety testing on the eve of the event.

Sam Altman opened the DevDay 2026 keynote by unveiling Dots, what OpenAI calls “remarkably capable, always-on agents built to handle everything.” Unlike Codex or ChatGPT, Dots are designed to operate independent of any single device or interface, pursuing user-defined goals continuously in the background with minimal oversight. OpenAI demoed Dots designing websites, building slide decks, and managing calendars.

Dots are available starting Tuesday in ChatGPT for Pro and Business Premium users in eligible markets, launchable from either Codex or ChatGPT. “Today, you can start with your primary dot, give it a name, and make it your own,” OpenAI said — with teams of Dots working together as the longer-term vision. Users can message Dots through Slack, Teams and other organizational platforms, with text-message support coming soon.

For enterprises, “specialist Dots” can be provisioned with specific identities, credentials and tools through existing systems, and OpenAI is already working with Microsoft to integrate into Agent 365 security controls. Much of the underlying capability existed in Codex-style agentic harnesses — Dots repackages it as a standalone always-on assistant, OpenAI’s answer to the persistent-agent wave that includes Meta’s Muse.

OpenAI's DevDay 2026 keynote goes live today, Tuesday September 29, at 10am PT / 1pm ET / 5pm UTC from Fort Mason in San Francisco, streaming free on OpenAI's site. In-person applications are closed, and OpenAI has published no product agenda — only the keynote is confirmed live, with other sessions recorded for later release.

The pre-event rumor mill is loud: TestingCatalog's Alexey Shabanov reports code references and a briefly-visible ChatGPT Pro page line pointing to 'o', an always-on assistant that keeps running multi-step tasks outside the active chat window, possibly launching across all three Pro tiers. A rumored $500/month 'ChatGPT Pro Max' plan, a GPT-6 Cyber preview, Managed Agents, and a wider Ultrafast API rollout are also in circulation — but none of it is confirmed.

The watch starts with OpenAI's own teaser: an X post from September 28 linking a video with no named product. The practical question for the keynote is what users can actually access, when, at what cost, and under what limits — the Wire will confirm or debunk the pre-launch rumors as the news lands.

OpenAI added hosted computer use to its Agents API at DevDay 2026, letting developers build agents that open a browser, inspect pages and work through website flows on their own, powered by GPT-6 Astra. The update brings Codex's multi-agent tools, tool search, tool calling and context compaction into developers' own products; it is available through the API and in Codex and ChatGPT Work on Pro 500 and Enterprise plans.

New alongside it is the Decisions API, a limited-preview service built on GPT-6 Luna for fast, low-latency single-choice tasks: routing requests, classifying images (visual inputs are accepted), or choosing an agent's next action from a predefined set of answers. It is not a replacement for open-ended report writing — its value is a small, fast decision inside a larger system — with a broad release planned in the coming days.

OpenAI also announced Bedrock Managed Agents with Amazon, a limited preview that runs OpenAI agents entirely inside customers' AWS environments via Bedrock, and open-sourced the Codex harness, the software wrapper behind its dots agents that lets outsiders inspect how an agent uses tools and tracks tasks.