Now, the venture capital firm is committing an additional $1.7 billion to that effort in a bet against fears of an artificial intelligence bubble, Bloomberg News reported Monday (Jan. 19).
For Andreessen Horowitz — or a16z, to use its finance world nickname — “infrastructure” means any AI software targeted towards technical buyers and not consumers, the report added. These include things like foundational models, networking security and coding applications.
“Some of the most important companies of tomorrow will be infrastructure companies,” Raghu Raghuram, a managing partner at the venture firm and former VMware CEO, told Bloomberg.
The report adds that these bets have begun paying off. For example, Stripe recently agreed to purchase a16z-backed billing platform Metronome for $1 billion while Salesforce acquired Regrello, an AI provider for manufacturers.
Other deals include Meta’s purchase of AI audio company WaveForms, while AI coding startup Cursor raised a new round of financing in November valuing it at $29.3 billion. When Andreessen Horowitz first backed the company in 2024, it was worth $400 million.
Andreessen Horowitz co-founder Ben Horowitz warned that it’s too soon to make any judgments about the fund’s performance, which is usually judged in terms of decades. So far, however, this is “one of the best funds, like, I’ve ever seen,” he said.
The Bloomberg report contends that it’s not yet clear whether the firm’s success with the fund exemplifies an AI bubble, or defies it.
As PYMNTS CEO Karen Webster wrote last week, the debate about the reality of an AI bubble has obscured what was happening within companies that build projects and run operations.
“Inside those organizations, the question was never whether AI was interesting. It was whether it could be fully trusted,” she added. “For much of the year, the answer was no. AI was used as a helper, not a decision-maker. It summarized, suggested and analyzed, but humans remained firmly in control of outcomes.”
But a shift happened late in the year, not involving artificial intelligence itself but in how businesses judged their own readiness and that of the technology.
Research by PYMNTS Intelligence followed 60 chief product officers at billion-dollar companies throughout last year, and the data shows a shift. In August of 2025, almost all (98%) of these executives said they were not willing to give autonomous agents any meaningful authority.
But three months later, that position had softened, with the share of firms merely considering AI for core operations dropped from 52% to 30% between August and November.
For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.
