$253 billion set the global record for AI investment in 2024. Then Stanford measured $581 billion in 2025 - more than double in a single year (Stanford HAI, 2026). But the more significant number isn’t the investment total. It’s what that capital is doing to specific jobs and functions.
Marketing is now the profession most transformed by AI productivity gains. Stanford’s 2026 AI Index, released this month, confirms: marketing sees a 50% productivity improvement from AI - the highest of any measured business function.
$581 Billion and the Fastest Technology Adoption in History
The macro picture first.
Global AI investment hit $581 billion in 2025, up from $253 billion in 2024 (Stanford HAI, 2026). The US captured $344 billion of that - more than 23x China’s $12.4 billion. Global AI compute has grown 3.3x annually since 2022, a 30x total increase since 2021.
More striking than the investment number: generative AI reached 53% population adoption in just three years - faster than the personal computer, faster than the internet (Stanford HAI, 2026). No consumer technology in modern history has been adopted at this speed.
On GitHub: 5.58 million AI-related projects in 2025, up 23.7% year-over-year and 5x from 2020 (IEEE Spectrum, 2026). Developers are voting with their code.
Marketing +50%, Software Dev +26%: Why the Uneven Distribution?
This is the most underreported data point in the report.
Stanford measured actual productivity outcomes by function after AI deployment. The results:
- Marketing and content creation: +50%
- Software development: +26%
- Customer support: +14 to +15%
The disparity isn’t random. Marketing is fundamentally language work - writing, editing, reframing, generating ideas. AI is built for exactly this. Customer support is more complex because it requires per-customer context. Software development involves logical reasoning chains and debugging that still require significant human oversight.
The implication is direct: if you’re a marketer who hasn’t integrated AI into your workflow, you’re operating at 1x while your AI-augmented peers are at 1.5x.
This isn’t a future risk. It’s a gap opening up right now.
The Stanford Paradox: 59% Support AI, 52% Feel Nervous
The report captures an interesting tension in global public sentiment:
- 59% believe AI’s benefits outweigh its drawbacks - up from 55% in 2024 (Stanford HAI, 2026)
- Yet 52% say AI-powered products make them nervous
These aren’t contradictory. They reflect a more nuanced reality: users see tangible benefits but remain uncertain about what they can’t see yet.
Trust in government AI regulation varies dramatically by country. Singapore leads at 81%. The United States - which invests the most in AI by far - scores just 31% (Stanford HAI, 2026). The countries building the most AI trust their regulatory capacity the least. That gap is worth watching closely.
Vietnam and Southeast Asia: 48x More Users, But Who Benefits?
One data point rarely surfaced in global AI coverage: Vietnam recorded a 48x increase in AI user numbers between 2023 and 2024 - among the fastest adoption surges globally (InvestVietnam, 2026). Currently 42% of Vietnam’s general population and 65% of SMEs are using AI tools.
Across Southeast Asia: the AI market at $12 billion in 2026 is projected to reach $80 billion by 2031 - a 37% CAGR (Source of Asia, 2026).
But rapid adoption speed doesn’t equal depth of proficiency. Stanford’s report also flags something important for high-adoption markets: entry-level workers in software development and customer support are seeing distinct headcount decreases, while mid-to-senior positions have held steady or increased.
For Vietnam and Southeast Asia - where a large cohort of young professionals is entering tech and marketing careers - this is a meaningful signal. Not in the science-fiction sense. More specifically: the entry-level position that previously needed 5 people now needs 3 people who use AI well.
The answer isn’t to avoid AI tools. The answer is to understand where that +50% productivity gain is actually accumulating - and position yourself on the right side of that divide before the market adjusts around you.
NateCue's Take
The most interesting number in Stanford's 2026 AI Index isn't the $581 billion in global investment. It's the productivity gap between functions: marketing +50%, software development +26%, customer support +14-15%. This isn't random. Marketing is fundamentally a language, idea, and output job - exactly where LLMs excel. Coding involves logical reasoning chains that are harder to fully delegate. The practical implication: a 50% productivity gain doesn't mean every marketer gets 50% more productive. It means the marketers using AI well are doing 1.5x the work. Everyone else is standing still - falling behind relative to their AI-augmented peers. The question to ask isn't "will AI take my job?" It's: "is someone using AI already doing my job better than I am?"