Technology & Science
Enterprise AI Spend Triples to $37 B in 2025, Triggering a Wave of Corporate Deals and Guardrails
Menlo Ventures’ new data show enterprise AI outlays jumping from $11.5 billion to $37 billion in 2025, cementing AI as the fastest-growing software category in history.
Focusing Facts
- Menlo Ventures finds AI applications alone absorbed $19 billion of 2025 spend, lifting enterprise AI to 6 % of the global SaaS market in just three years.
- Days after the report, Anthropic inked a three-year pact with Accenture to embed its models in client workflows, joining 128 licensing deals and 31 AI agents already live at One NZ.
- One NZ says autonomous AI cut projected head-count needs by 20 % and trimmed network energy use by 20 %, while Meta began paying large publishers for real-time news feeds—deals smaller outlets can’t access.
Context
Spending curves like this last appeared when U.S. electrification outlays swelled seven-fold between 1900-1910, vaulting utilities from novelty to infrastructure. Today’s AI surge likewise reflects a structural shift: capital is pooling around model providers (Anthropic, OpenAI), integrators (Accenture) and early adopters (Nvidia-powered telcos), while content owners and regulators scramble for leverage. Copyright payments to select media echo 1880s railroad rebates that favored big shippers over small farmers, foreshadowing antitrust pressure. On a 100-year arc, the question is not whether AI becomes ubiquitous—like power grids or the Internet—but who captures the rents and whether late-adopting regions are locked out, as the UNDP warns. Tripling spend in twelve months signals a break-neck diffusion rate; history suggests such booms often overshoot before standards, labor policy and energy capacity catch up.
Perspectives
Business and investment-focused media
e.g., Yahoo! Finance, The Motley Fool, NZ Herald, The Manila Times, Entrepreneur — Present AI as an unprecedented economic catalyst and urge companies and investors to move fast—adopt the technology enterprise-wide and buy leading AI stocks before the next market surge. These outlets profit from market enthusiasm and advertising tied to finance and tech, so they may accentuate upside forecasts and gloss over labour displacement or regulatory costs highlighted elsewhere.
Public-sector and social-impact outlets
e.g., The Malta Independent Online, The Star/AP, edinburghlive local government coverage — Warn that AI could entrench inequality, distort democracy and requires tight human oversight, transparency and new policies to protect vulnerable communities. Their mission to safeguard the public may lead them to spotlight worst-case scenarios and underplay the productivity gains that businesses report.
Publisher-rights tech journalism
e.g., PCMag UK — Argues that AI platforms’ content deals enrich big media while starving smaller sites of traffic and revenue, urging payment and stricter limits on scraping. PCMag’s parent company is suing OpenAI over copyright, giving the outlet a direct financial stake that can colour its coverage toward emphasizing harms and legal jeopardy for AI firms.