2026.11.10 AI News Briefing — Anthropic's IPO Roadshow and an AI Training Opt-Out Signal
Contents
Anthropic kicked off its Nasdaq IPO roadshow, while Meta paused its publisher content-crawling terms. Four Big Tech companies agreed on a shared AI training opt-out signal, and in Korea, Nvidia's AI Day Seoul and Kakao's new open-source model stood out.
1. Anthropic Kicks Off Nasdaq IPO Roadshow in Earnest
- Source: LetsDataScience | S-1 filed Oct. 26, 2026; roadshow began the week of Nov. 9
- Key details: Anthropic filed its S-1 for a Nasdaq listing (ticker "ANT") on the 26th of last month. It disclosed first-half revenue of $16.3 billion, a net loss of roughly $80.7 billion, and $141.1 billion raised to date. The founders' LLC holds 50.1% of the voting power, and a Long-Term Benefit Trust (LTBT) elects a majority of the board. The offering price and share count remain undisclosed.
- Why it matters: Amid an IPO race with OpenAI, this is the first time an AI company's actual revenue and loss figures have been disclosed at this scale.
2. Meta Indefinitely Delays Enforcement of Publisher-Crawling Terms
- Source: Niemanlab | Nov. 5, 2026
- Key details: Meta has indefinitely delayed the revised commercial, catalog, and generative-AI ad terms that were set to take effect on November 20. The backdrop is pushback from the News/Media Alliance, which represents roughly 2,000 outlets, and a UK Competition and Markets Authority (CMA) investigation opened on October 30. The existing terms remain in place.
- Why it matters: The platform-versus-press conflict over whether AI training data collection should be "opt-out by default" or "subject to separate agreements" looks set to continue.
3. Google, Microsoft, OpenAI, and Anthropic Adopt a Shared AI Training Opt-Out Signal
- Source: PPC Land | Oct. 30, 2026
- Key details: The four companies agreed on a signal, based on an IETF draft, that lets sites opt out of AI training with a single shared token in robots.txt. Rollout timing varies: Anthropic immediately, OpenAI within a few months, Microsoft by February 2027, and Google targeting 2027. The signal does not apply to training datasets already collected.
- Why it matters: This is the first move toward standardizing AI crawler opt-out rules that had previously varied from site to site.
4. Kakao Unveils Multimodal Decision Model 'Kanana-2-V-Decide'
- Source: Korean Center | Nov. 6, 2026
- Key details: Kakao open-sourced a decision model on Hugging Face, built on its own multimodal model "Kanana-2-V-3B," that looks at text and images together and outputs only probabilities or scores. It said the model can be used for the "judgment" step of agent pipelines, such as identifying photos of damaged products or auto-classifying customer inquiries, with an average response time under 0.04 seconds on a single H200.
- Why it matters: Unlike general-purpose LLMs that generate long responses, a lightweight model that outputs instant judgments only is a practical way to cut the cost and latency burden of agent pipelines.
5. Nvidia's 'AI Day Seoul 2026' Opens at COEX
- Source: CEOScoreDaily | Nov. 9, 2026
- Key details: Nvidia held 'AI Day Seoul 2026' with 20 Korean partners at COEX in Seoul on the 9th and 10th. Samsung Electronics, SK Hynix, SK Telecom, LG Electronics, Naver Cloud, and others took part in 15 sessions on agentic AI, physical AI, and AI infrastructure, with roughly 2,000 attendees expected.
- Why it matters: It's a signal that Nvidia sees Korea as a key partner market for the agentic and physical AI ecosystem.
Today's Takeaways
- Big Tech is simultaneously pursuing IPO processes that prove "money made from AI" and processes that take responsibility for "AI training data" — pausing terms, adopting shared opt-outs.
- In Korea, the emphasis is less on large-model competition and more on lightweight, specialized models that plug directly into agents (Kakao) and infrastructure-ecosystem events (Nvidia).
References
Contents
Related posts
2026.11.07 AI news briefing — personal finance AI and Korea's advertising AI race
Claude expands into personal finance, while Naver and Kakao roll out AI advertising assistants. Also covered: OpenAI's funding round and Nvidia's Seoul event.
2026.11.03 AI news briefing — funding squeeze and a monetization test
Five key AI stories from Korea and abroad today: OpenAI's funding cutback, diverging earnings outlooks for Naver and Kakao, and Nvidia's AI Day Seoul.
2026.11.09 AI News Briefing — AI valuations wobble as Korean agent competition heats up
As OpenAI's and Anthropic's valuations are repriced in opposite directions amid market turmoil, Naver and Kakao are moving up their AI agent launches in Korea, and the AI chip shortage is spreading into the auto industry.
2026.11.08 AI News Briefing — Data Center Accountability and AI Commercialization in Korea
Today's five AI stories: OpenAI's data center cost pledge and scaled-back funding round, Anthropic's new finance feature, and Naver and Kakao's AI advertising assistants.