Agents push enterprise AI spending toward a fivefold surge

Companies Are Spending 5x More on AI Now

The shift from seat-based software to token- and compute-driven agents could reshape budgets, vendors, and the economics of enterprise technology.

3 key takeaways

1Enterprise AI spending is accelerating as agents turn software usage into token- and compute-based consumption.
2Anthropic leads enterprise API spending, while coding agents drive rapid growth in measurable, high-value workflows.
3AI-native companies are challenging SaaS incumbents, but most current AI budgets still supplement rather than replace existing software.

Don't miss

The episode’s clearest inflection point is the finding that enterprise AI budgets are shifting from seats to tokens and compute as agents perform more work.

The brief

The episode opens against claims that AI adoption is slowing, arguing that enterprise demand is accelerating as agents move AI from assistants toward autonomous work.

Menlo Ventures’ enterprise research points to a spending surge driven by tokens and compute rather than seats, with companies routinely spending beyond traditional software budgets.

Frontier models still command enterprise demand despite cost concerns, while Anthropic has gained API share through Opus and agentic coding as OpenAI rebuilds momentum.

Coding agents and financial-services applications are growing fastest, suggesting that AI adoption advances most quickly where results can be measured or verified.

The central tension is whether AI spending will expand software markets or eventually displace SaaS, as AI-native firms attack new categories while incumbents defend embedded workflows.

What was said on this episode

16 statements · 7 positive · 7 negative · 2 neutral

Nathaniel Whittemoreon Enterprise AIPositive0:28
Enterprise AI adoption growth is accelerating.
“The speed at which it is growing is actually increasing”
Listen at 0:28
Nathaniel Whittemoreon AI stocksNegative1:48
AI stocks experienced their worst drawdown since 2020.
“AI stocks plummeted in their worst drawdown since 2020”
Listen at 1:48
Nathaniel Whittemoreon Nvidia CorporationPositive2:49
Nvidia traded at a 13x multiple and appeared among the market’s cheapest stocks.
“with Nvidia trading at a 13x multiple, making it one of the quote-unquote cheapest stocks on the market”
Listen at 2:49
Nathaniel Whittemoreon AI bull marketNegative4:16
The hedge-fund deleveraging could end the AI bull market.
“that could mark the end of the AI bull market”
Listen at 4:16
Nathaniel Whittemoreon Anthropic IPONegative4:39
Prediction markets increasingly price an Anthropic IPO delay or cancellation.
“prediction markets are beginning to price in a delay or even a cancellation”
Listen at 4:39
Nathaniel Whittemoreon Meter independent auditor suiteNegative8:34
Using only Meter creates an incomplete independent AI auditor suite.
“having only Meter here means a very incomplete independent auditor suite”
Listen at 8:34
Nathaniel Whittemoreon ClaudeNegative8:54
Claude does not provide investment advice in its personal-finance product.
“Claude won't even give investment advice”
Listen at 8:54
Nathaniel Whittemoreon MetaMuseNegative9:53
Meta’s Muse avatars cannot use visual inputs to detect nonverbal cues.
“they don't actually have visual inputs to pick up on non-verbal cues”
Listen at 9:53
Nathaniel Whittemoreon Enterprise AI spending modelNeutral16:18
Enterprise AI spending shifted from seat-based software toward token-based usage.
“AI spend stopped being about seats and started being about tokens”
Listen at 16:18
Nathaniel Whittemoreon AI agentsNeutral17:10
Agent spending scales with workload rather than employee headcount.
“Spending then scales with work, not with headcount”
Listen at 17:10
Nathaniel Whittemoreon Smaller efficient AI modelsPositive21:37
Smaller efficient models will increasingly join enterprise AI stacks.
“you will see those smaller, more efficient models start to come in and form a more complex stack”
Listen at 21:37
Nathaniel Whittemoreon AnthropicPositive22:05
Anthropic continues gaining enterprise API market share.
“Anthropic has continued its ascendancy”
Listen at 22:05
Nathaniel Whittemoreon OpenAI enterprise strategyPositive23:03
OpenAI’s enterprise focus is producing positive results.
“OpenAI's emphasis throughout 2026 on enterprise and business use cases is clearly bearing fruit”
Listen at 23:03
Nathaniel Whittemoreon Anthropic and OpenAI enterprise AI spendingPositive23:49
Anthropic and OpenAI dominate enterprise AI spending competition.
“This is a completely two-horse race between Anthropic and OpenAI”
Listen at 23:49
Nathaniel Whittemoreon Open-source AI token sharePositive24:35
Token share will matter more than spending share over the medium to long term.
“in the medium to long term, the token share numbers are more significant than the spending numbers”
Listen at 24:35
Nathaniel Whittemoreon SaaS disruption from AI agentsNegative29:57
Agents using software and headless software pose a future SaaS disruption risk.
“the rise of agents using software and headless software”
Listen at 29:57

Statements are attributed to the speaker as said on the episode and reflect their view at the time, not PodLume's. They are not advice.

Weekly email

State of the Pod, every Monday

The week's most-discussed names and stories from podcasts, with the moments behind them. Free, five minutes, unsubscribe in one click.