SecondSourceAI Industry Insight · Full Archive

Daily Brief SecondSource Morning Brief · October 6, 2026 · Oct 6, 2026

How we checked each item today

For readers who want to dig in. If you only want what happened, go back to Today.

This issue draws on our internal research brief from early October 6 and last night's material. All three main-line items were published or happened on October 5; item 1 also cites background sources from September 22 and September 30. The two unverified items are October 5 news, and the archive pick comes from our September 19 deep dive. We swept 431 pieces overnight, and this issue uses 9 outside sources with links you can check.

1. [Today] (article October 5) SemiAnalysis tests same-price US$200 subscriptions: at API list prices, the usage Claude's mid-tier model gets is worth about five times ChatGPT's

How we checked

the five-times figure and the half-allowance cap come from SemiAnalysis alone, measured on a small number of accounts; we read the public sections, not what sits behind the paywall. The OpenAI allowance cut is relayed by WinBuzzer; we didn't read OpenAI's own plan page. ⚠️ The five-times figure prices usage at list rates; it does not mean five times as much work gets done: different models burn different amounts of usage on the same task, and list prices themselves distort the conversion, as the authors acknowledge; the article is also selling the firm's own paid dashboard. ⚠️ Our analysis was produced with help from Anthropic's models, and Anthropic is a party to this item.

Back to this item ↑

2. [Today] (announced October 5) OpenAI says it will add invisible watermarks to ChatGPT and Codex text output in the EU in the coming weeks; the detection tool opens first only to approved researchers

How we checked

we read OpenAI's announcement itself, not the technical report. Every number is OpenAI's own measurement with no third-party check; its claim that performance "matched or exceeded" SynthID, Google's text watermark, is also self-assessed.

Back to this item ↑

3. [Today] (hearing October 5) New York City Council's AI hearing is done: representatives of Anthropic, OpenAI, Google and Meta all testified by video, SpaceX was subpoenaed and didn't appear, and there was no vote

How we checked

For CBS, we read the key excerpts on the page, not the full report. Of CNBC's same-day report we read only the headline; a search summary mentioned an Anthropic whistleblower testifying at the hearing, but we didn't read the original and don't use it. ⚠️ In the passages we read, no company stated support for or opposition to the third-party validation bill above.

Back to this item ↑

What you are not getting today. What would most affect judgment is a second, independent measurement of the "about five times on same-price plans" figure; we don't have one. Beyond that, we don't have the paywalled sections of SemiAnalysis, OpenAI's plan page, the textGrain technical report, the full hearing record and CNBC's report, or the Bloomberg original.

Sources & accounting

The past 24 hours. 431 new pieces came in from October 5 to 6: 110 social-platform posts, 269 academic papers, 9 other papers, 14 show transcripts, 8 subscription newsletters, 1 industry analysis and 20 company and personal blog posts. Of those 431, we read 27, filtered out 12 and left 392 unread, mostly 268 academic papers and 86 social posts. We haven't yet read the same day's newsletters from Zvi Mowshowitz, Import AI, Stratechery and Understanding AI, and today's judgments don't use them. Most of the 9 outside sources cited in this issue's body are pages and reports we checked or added separately, so not all of them are among those 431.

One-time backfill. No new one-time backfill today.

A note on source concentration. ⚠️ The five-times figure and the half-allowance cap in main-line item 1 come from SemiAnalysis alone. ⚠️ Every number in main-line item 2 is OpenAI's own measurement. ⚠️ Main-line item 3 rests on a single on-the-scene report from CBS. ⚠️ Our analysis was produced with help from Anthropic's models, and main-line items 1 and 3 both involve Anthropic directly.

The sources we track. Our long-term roster has 529 named sources: 302 on social platforms, 90 shows, 51 news outlets, 48 blogs, 48 paper authors and 46 newsletters, with the rest spread across earnings, keynotes and other channels. ⚠️ Those are counts of tracked sources, a different population from the 431 new pieces above. Representative names: on social platforms, Ben Horowitz, Ethan Mollick, Eric Topol and Teortaxes; among newsletters, SemiAnalysis, Zvi Mowshowitz, Import AI and Understanding AI; among shows, Cognitive Revolution and Latent Space. This issue uses 9 outside sources in the body. We count only links the body actually cites that are not on our own domain, so it can't be compared with the 529 tracked sources either. Taiwanese sources: 7 new in September, 119 in total; Taiwanese media, Taiwanese companies' official channels and local Taiwanese reporters count, foreign outlets relaying Taiwanese media do not.