Happy Friday. Anthropic wants to go public before Thanksgiving, and prospective investors are pricing it at up to $2 trillion. That would make it the largest IPO in history and more than double the company's estimated valuation from just five months ago.
Elsewhere, Salesforce is paying a reported $2 billion for a two-year-old customer research startup with $30 million in annualized revenue, and Microsoft is undercutting voice AI vendors on price while claiming the top accuracy benchmark. The AI spending race is accelerating on every front.
FUNDING & M&A
Anthropic Eyes History-Making IPO Before Thanksgiving
Anthropic plans to begin its IPO roadshow the week of Nov. 9 and start trading before Nov. 26. Revenue hit $11.5 billion in Q2 2026 alone, but operating losses reached $8.06 billion in 2025, and the company faces at least $518 billion in infrastructure commitments over the next decade, roughly 80% of which are non-cancelable.
The filing's risk section is the story within the story: Anthropic spent more pages warning investors about AI catastrophe than describing its own business. One company safety researcher put the odds of AI killing humans within a decade at above 10%. For founders watching this IPO, that disclosure strategy is worth studying. It may deflect future regulatory and legal risk, or it may prove to be the strangest prospectus language in public market history. Either way, the roadshow will set a valuation benchmark that shapes how every AI company is priced for years.
FUNDING & M&A
Salesforce Pays 67x Revenue for Two-Year-Old AI Research Startup
Salesforce signed a definitive agreement to acquire Listen Labs, founded in 2023, for a reported $2 billion. The startup generates about $30 million in annualized revenue, grown 15 times over nine months, and ran AI-moderated interviews for customers including Microsoft, Canva, and Anthropic before the deal was announced.
Listen Labs walked away from a signed $125 million Series C term sheet to pursue this sale, a move sources called rare and generally frowned upon in venture circles. That detail matters as much as the price. At 67 times revenue, this multiple will fuel expectations across the customer research and voice analytics space. Founders in adjacent categories should note what Salesforce is actually buying: a recruiting network of 50 million people and the ability to simulate customer behavior before a product decision is made. That capability slots directly into Marketing Cloud and Service Cloud in ways that pure survey tooling never could.
ARTIFICIAL INTELLIGENCE
Microsoft Tops Accuracy Benchmark, Undercuts ElevenLabs on Voice AI Price
Microsoft released MAI-Transcribe-2-Streaming on Oct. 1, a model that begins producing transcript results in just over 100 milliseconds and confirms a final transcript 0.13 seconds after a speaker stops. It ranked first among 38 models on the Artificial Analysis streaming benchmark with a 2.5% word error rate. The introductory price is $0.54 per audio hour through the end of 2026.
Microsoft is pricing this to take share from dedicated vendors. Its streaming rate converts to roughly $9 per 1,000 minutes against $6.50 per 1,000 minutes for ElevenLabs and Deepgram, so it is not the cheapest option yet. But Mustafa Suleyman's claim of 55% faster and 60% cheaper than ElevenLabs applies to a different comparison point, and Microsoft's introductory pricing will likely move. Developers building voice agents who are currently locked into third-party transcription vendors now have a credible enterprise alternative inside the same platform where many are already running inference and storage.
Created by the robots at The SaaS Sentinel