Corporate research and development spending among the world's largest companies reached a record USD 1.5 trillion in 2025, rising nearly 8 percent and climbing to 6 percent of revenue, the highest share ever recorded. The findings come from the Global Innovation Index 2026 Tracker, published by the World Intellectual Property Organization (WIPO), which charts innovation trends through a year dominated by artificial intelligence. The report finds that innovation has held up across nearly all measures, marking a broad improvement from the mixed picture presented in recent years, though recovery remains incomplete with about three-quarters of indicators still below their decade averages.
Global R&D intensity rose from 1.4 percent of GDP in 2000 to close to 2 percent in 2025 and is projected to hold there through 2026, representing the highest research effort in history. Scientific publications reached a record 2.4 million articles in 2025, up 6.9 percent from 2024, led by China's 15 percent increase, which now accounts for 37 percent of the global total. Venture capital deal values surged 27.9 percent to USD 510 billion, the highest since 2021, even as deal counts fell 1.4 percent to about 43,500, reflecting capital pooling into fewer, higher-value firms. High-tech exports grew roughly 14 percent to a new high, nearly three times faster than global trade, driven by AI infrastructure build-out. Labor productivity worldwide rose 2.7 percent, above the decade average of 2.2 percent, with workers now generating close to USD 57,000 in output per hour. Three indicators worsened: VC deal counts continued falling for a fourth consecutive year, renewable energy costs rose for the first time since 2014, and global temperatures were the second highest on record at 1.19°C above the 1951-1980 baseline.
The report identifies two defining forces that pull in opposite directions. According to the authors, "the first is concentration at the frontier," with a growing share of investment directed toward AI and a narrower set of firms and locations, while "the second is diversification across the base," as innovation and technology spread to middle-income economies beyond China. AI now accounts for about half of global VC value, up from roughly 27 percent in late 2022, with Northern America capturing about USD 213 billion of the USD 270 billion in AI-related VC in 2025. Within corporate R&D, software and AI-intensive firms grew fastest, with Meta expanding research spending by around 32 percent and NVIDIA by about 43 percent. The 125 new unicorns that emerged in 2025 represented the highest count since 2022, with OpenAI, Anthropic and xAI recording the largest deals globally in early 2026, together representing close to 43 percent of total VC value in the first half of the year.
Why the split matters: the report explains that optimism around AI fueled large capital spending that supported global growth and kept financial conditions accommodative in 2025, but this concentration leaves the recovery vulnerable if expected AI returns fail to materialize. Meanwhile, the diversifying base shows innovation spreading to unexpected places: Egypt, Chile and Ethiopia raised scientific output; Malaysia, Indonesia and Viet Nam sustained long-run growth in R&D; VC and patenting activity grew from Argentina to Morocco, India to Türkiye; and labor productivity rose fastest across China and a diverse group of economies in Asia and Africa. The global stock of industrial robots reached nearly 4.7 million units in 2024, up about 9 percent, with adoption widening beyond leaders as India's stock grew 17 percent and Viet Nam's by 10 percent. Electric vehicle stocks rose 629 percent in the Philippines, 155 percent in Viet Nam, and 121 percent in Türkiye between 2024 and 2025, though from low starting points.
The near-term environment is more demanding than a year ago, with global growth expected to ease from about 3.4 percent in 2025 to around 3 percent in 2026 before recovering modestly in 2027. The report concludes that the open question remains whether new innovation waves can revive sluggish productivity growth, with AI now the leading candidate, though its contribution to aggregate productivity isn't yet visible and its lasting effect is likely to come from a second wave of investment in data, skills and reorganized processes rather than from physical infrastructure alone. The largest hyper-scalers are set to spend well over USD 1 trillion on AI infrastructure across 2025 and 2026, increasingly financed through debt and private credit, meaning a downward revision of expected AI returns could tighten financing conditions well beyond the sector. Whether the concentrating frontier or the diversifying base comes to dominate will shape both growth and the geography of innovation in the years ahead.