
Welcome to the Second Opinion monthly update. Today we update on secondary investors’ pre-midterm sentiment, then walk through the largest AI ideas, beyond foundation models and neoclouds, that the private market is likely to rethink.
The midterms looked in part like a referendum on the AI buildout. In a Gallup study, only 36% of Americans expect AI to mostly help their country, against 93% in China, which put the US near the bottom of the countries surveyed.
In 2026 Americans’ mood about AI got worse very quickly. “It’s very rare … to see voters move so negatively on an issue so quickly,” McKenzie Wilson, a messaging strategist at the progressive firm Blue Rose Research, told POLITICO in late October.
At the same time, businesses were quick in adopting AI. AI is still the main story of business growth (which voters eventually need). In September, 55.9% of US businesses on Ramp paid for AI, up 12.1 points from a year earlier (Ramp AI Index).
Whoever wants these voters will need to rebrand AI at the least, and probably reshape parts of the market (and we may never learn whether AI was a bubble that would have deflated on its own). 🙂
The private market is “in talks”, the secondary market is waiting
It’s fair to say the private market in October was overly optimistic, and some of its loudest voices now admit it. October set up expectations of a new cohort of $20B companies, mostly AI.
The rounds below were closed or in talks during October, sorted by valuation.

Valuation and status as reported, October 2026. Source: LBX Pro.
Secondary investors didn’t show much optimism in October. They are waiting.
PitchBook’s latest US VC Secondary Market Watch (October 28) says the venture secondary market has just passed its peak, with volume down 9% in Q3 versus Q2. PitchBook ties the peak to SpaceX’s IPO and expects a larger pullback once Anthropic and OpenAI exit (PitchBook).
LBX Pro data shows the same thing from the order book: the market is ask-heavy.

Source: LBX Pro listed orders (not trades), September–October 2026.
Sellers are trying to sell while buyers hold back with record dry powder. PitchBook expects VC secondary dry powder to reach $14.6 billion in 2026 as new buyers enter.
In moments like this, we analysts usually say the market will be repricing. Yes, but this time it is also waiting for a bunch of AI ideas and moats to be justified, and not just to investors but to the wider public.
When tech was the good guy
Tech has had public approval before, and it lost it within a few years. In 2010, 68% of Americans told Pew that tech companies had a positive effect on the country. That peaked at 71% in 2015, when only 17% saw a negative effect (Pew Research Center, Poets&Quants summary of the 2010 figure).
That was the era of Google’s “Don’t be evil” and of Facebook being credited with helping organize the 2011 Arab Spring. Tech was seen as a public good, not just a profitable industry.
By 2019, after fights over privacy, misinformation and harassment, positive views had fallen to 50% and negative views had nearly doubled to 33%. The drop happened in both parties.
And now winning back the 2015-style consensus may matter as much as the next revenue milestones.
With that in mind, here is what the market has in line, how secondaries price it, and who is most exposed to repricing.
How the market prices the next wave
Start with the two giants as a benchmark. At a $1.25 trillion IPO valuation and a run rate above $80 billion, Anthropic would trade at about 15.6x annualized revenue (AIStockWire, citing WSJ). OpenAI is raising at $1.25 trillion on roughly $50 billion, or 25x, a private-market premium.
The next wave is priced far higher.

Multiple = implied valuation as of 1 November 2026 ÷ annualized revenue (ARR or run rate, month shown). Exposure is LBX Pro’s editorial rating. Sources: LBX Pro implied valuations; ARR statements from LBX Pro Signals.
What each company is betting on, and what the midterms change for it:
- ClickHouse: fast, cheap database becoming the default store for AI and agent data. After the midterms: runs on existing cloud capacity, so no direct hit.
- Cognition: AI software engineers (Devin + Windsurf) that replace engineering work. After the midterms: most exposed to “AI takes jobs” politics; heavy inference costs.
- Sierra: AI customer-service agents paid per resolved conversation. After the midterms: replacing call-center jobs is an easy political target; uses little compute.
- Legora: AI platform for lawyers winning global enterprises (Salesforce). After the midterms: Swedish and global.
- OpenEvidence: free clinical AI used by ~⅔ of US physicians, paid for by pharma ads. After the midterms: Democrats focus on pharma marketing; states are passing health-AI laws.
- Fireworks AI: fast, cheap inference for open models. After the midterms: rents GPUs, so dearer power and capacity hit margins.
- ElevenLabs: leading AI voice platform; 55%+ enterprise revenue; India is the #2 market. After the midterms: voice cloning is a bipartisan deepfake target (NO FAKES Act, state laws); global revenue softens it.
- Ramp: profitable corporate card becoming a full finance stack. After the midterms: not an AI-infrastructure bet.
- Databricks: enterprise data + AI platform growing. After the midterms: diversified.
- Lovable: “vibe-coding” app builder used by ⅔ of the Fortune 500. After the midterms: costlier compute raises per-token costs; hobby users may churn.
- Fal: fast inference for AI images, video and audio. After the midterms: needs a lot of GPUs; also exposed to deepfake and copyright rules.
The lesson? We need a few more metrics when we look at a company. Valuation growth on the primary market – sure. ARR growth – absolutely. But it’s time to watch how voters feel about AI ideas more closely too. Luckily, AI gives us enough data to do just that.
Victoria Arinstein
Product & Intelligence, LBXpro
More insights
All insights
Anthropic goes public. What will secondary money make?
Anthropic starts trading today. How secondary buyers did since 2024, why most bids came late, and what revenue forecasts and AI timelines miss.
Read
Half the Market Is on Strike
September's IPO delays, record private rounds, and a secondary market where only 1 in 5 names has buyers and sellers. Our Q3 data, explained.
Read
Is the market immune to good news?
Nvidia posted its best quarter ever, yet AI sentiment barely moved. What August's LBX25 Index and private-market data say about where AI risk sits.
Read