On mvanemmerik.com, Nov 16: NFL Week 10 chalk talk: Detroit dazzles, Cincinnati takes the North
vanemmerik.aiClaude, ChatGPT and Grok news, twice a day

Friday, November 6, 2026

Investors reportedly push Anthropic to lower its $1.5 trillion IPO valuation after a selloff

Reports say big prospective buyers want a lower price, a smaller raise or a later date for Anthropic's IPO; Anthropic hasn't changed its timetable. OpenAI reportedly banked the first $11 billion of its own round at $1.25 trillion. METR says Anthropic's progress has about doubled in pace, researchers detailed a Claude Dispatch flaw reported in May, and Claude Enterprise added IP allowlists and spend tags. Codex CLI 0.167 shipped. Quiet for xAI.

Friday, November 6, 2026

OpenAIBusiness

Update: OpenAI reportedly banks the first $11 billion of its new round at a $1.25 trillion valuationReported

Bloomberg reported on Nov. 6, citing people familiar with the matter, that OpenAI received the first $11 billion of its latest funding round at a $1.25 trillion valuation, and that the round could grow to about $20 billion in the coming weeks. OpenAI had originally sought $30 billion or more at a $1.4 trillion valuation before scaling back during the AI stock selloff. OpenAI had not commented on the report.

Why it matters: It sets a price for OpenAI well above its $852 billion March valuation but below its first target, a data point for investors weighing Anthropic's IPO.

OpenAIHappened Nov 6, 2026 · Sources: Investing.com (summarizing Bloomberg) · Reuters via TradingView

ClaudeFeature

Claude Enterprise adds project tags in beta to track spend by charge code or client-matter number

Eligible Claude Enterprise organizations can now tag projects with codes they already use, such as charge codes or client-matter numbers, and view most of the spend inside those projects by tag. Owners create the tags; spend appears under Usage in organization settings and can be exported. Some setups, such as HIPAA or zero-data-retention organizations, are excluded, and the Claude API and Console aren't covered.

Why it matters: Law firms, consultancies and other businesses that bill clients by code can now see what Claude costs per client or project.

ClaudeHappened Nov 6, 2026 · Sources: Claude release notes · Data Studios

OpenAIFeature

Codex CLI 0.167 adds an agent command center that marks unread results and a Vertex AI setup wizard

OpenAI shipped Codex CLI 0.167.0 on Nov. 6. codex agents --no-daemon opens an agent command center that flags unread results, the session picker can switch between connected backends without losing drafts, and an opt-in gcp_setup_wizard sets up Google Cloud Vertex AI. Saved chats now stay visible after switching model providers.

Why it matters: It continues OpenAI's push to make Codex work across clouds and juggle several agents at once from the terminal.

Try it: Update Codex CLI and run codex agents --no-daemon.

CodexHappened Nov 6, 2026 · Sources: ChatGPT & Codex changelog

ClaudeSafety

Researchers say a browser extension could use Claude Dispatch to run code on a linked computer, a flaw reported in MayReported

Forever Security researcher Gal Weizman published "ThisBash" on Nov. 4: a browser extension with access to claude.ai could reach a signed-in user's Dispatch session, send prompts to Claude Desktop and auto-click approvals, getting it to install and run an npm package, even on another computer paired to the same account. The researchers say Anthropic confirmed the report in May but closed it as a duplicate of a known issue, and on Oct. 13 said Dispatch would be retired by the end of October. Anthropic has not commented publicly on the write-up, and coverage could not confirm Dispatch is fully gone.

Why it matters: Agents that act across devices make the browser a path to the desktop, so extension hygiene now matters for AI accounts too.

Try it: If you still have Dispatch or remote control features enabled, turn them off where you don't need them, and review which browser extensions can access claude.ai.

ClaudeHappened Nov 4, 2026 · Sources: Forever Security · Cyber Security News

ClaudeBusiness

Update: Investors reportedly push Anthropic to lower its $1.5 trillion IPO valuation after a market selloffReported

The Wall Street Journal reports that three major investors planning to join Anthropic's IPO think the proposed $1.5 trillion valuation may need to fall, with one suggesting closer to $1 trillion; another floated raising about $80 billion instead of up to $100 billion, and one raised a delay. Reuters separately reports investors doubt the roadshow can start next week. Anthropic has not announced any change to its timetable and had not commented.

Why it matters: The listing was planned for the week of Nov. 9 with trading on Nasdaq under ANT before Thanksgiving, so any repricing or delay would reshape the year's biggest IPO.

AnthropicHappened Nov 5, 2026 · Sources: Business Standard (summarizing WSJ and Reuters)

ClaudeSafety

METR says Anthropic's model progress has roughly doubled in pace, most likely because AI is speeding up its researchReported

The nonprofit evaluator METR, which Anthropic invited to review internal data and interview staff over about 11 weeks, estimates Anthropic is now improving its models on well-specified tasks about twice as fast as before, and says growth in training compute can't explain it. METR cautions the speedup may be smaller on messy, hard-to-grade work, and says Anthropic's redactions limited what it could publish. This is METR's assessment, not an Anthropic announcement; Anthropic did not immediately comment.

Why it matters: It is a rare outside check on whether a frontier lab's own AI is accelerating its research, a threshold Anthropic's IPO filing flagged this week.

AnthropicClaudeHappened Nov 5, 2026 · Sources: Fast Company (AP)

ClaudeFeature

Claude Platform and Claude Enterprise add an IP allowlist in beta, blocking requests from unlisted addresses

Admins can now list up to 1,000 public IPv4 or IPv6 addresses or ranges, and requests from anywhere else are refused: on the Claude Platform that covers the Console and core API endpoints, and on Claude Enterprise it covers the Claude apps, Claude Code with a Claude sign-in, Cowork and more. It does not apply to Claude on Amazon Bedrock, Google Cloud or Microsoft Foundry. The same release adds a beta Admin API call that signs an Enterprise member out everywhere.

Why it matters: It limits where a leaked API key or stolen sign-in can be used, a common ask from security teams.

Try it: Platform admins: Claude Console > Settings > Identity and access; Enterprise owners: Organization settings on claude.ai. Tick the option that lets admins reach the page from any IP to avoid locking yourself out.

Claude APIClaudeHappened Nov 5, 2026 · Sources: Claude Platform release notes · Claude docs: IP allowlisting

OpenAIBusiness

Greg Brockman says OpenAI could be ready for an IPO in 2027 as enterprise becomes most of its revenueReported

Speaking at a Wall Street Journal tech conference, OpenAI president Greg Brockman said he thinks the company could be IPO-ready next year, according to TechCrunch. He said annualized revenue grew 70% in the third quarter from the second as enterprise revenue doubled, so enterprise is now the majority of revenue, and admitted OpenAI had lagged in agentic coding early in the year. OpenAI has not set an IPO date.

Why it matters: It is the clearest signal yet on OpenAI's listing timing, landing as Anthropic's own IPO meets investor pushback.

OpenAIHappened Nov 4, 2026 · Sources: TechCrunch