SILICON VALLEY -- Artificial intelligence startups raised $5.02 billion worldwide in 2016 to hit a five-year high, as application of the technology expands from online services and finance to such far-ranging fields as agriculture.
This trend is likely to continue through 2017, as investors scout out promising technologies and talent.
AI-related startup funding rounds totaled 658 last year, up from 493 rounds in 2015, when funding reached $3.12 billion, according to U.S.-based research firm CB Insights.
So-called mega-rounds are also increasing. For example, Texas-based StackPath, a startup offering AI-based cybersecurity services, took in $180 million. California-based Zymergen, a biotech venture that aims to automatize microbe engineering, raised $130 million.
Investors are also looking outside the U.S. Artificial intelligence-related fundraising by non-U.S. startups grew to 38% of the global total in 2016, up from 21% in 2012. The field is diversifying, with capital going to new players from countries such as the U.K., Canada, France, Israel, China and India.
Japanese AI startups also secured money last year, such as LeapMind, which creates services that use a kind of machine learning called deep learning, and fellow Tokyo venture Xenodata Lab, a developer of financial analysis tools for retail investors.
AI funding is coming not only from major Silicon Valley venture capital firms such as Khosla Ventures and Andreesen Horowitz, but also from Intel, Google, General Electric and other big companies. Microsoft set up a special AI investment fund in December.
Advances in machine learning algorithms will continue to broaden the applications of AI, says CB Insights analyst Deepashri Varadharajan, who predicts that investment will continue flowing into the field.
But some are cautious, since AI has experienced two burst bubbles in the past, leading to a period when investor interest cooled.
There are some bad signs, such as the growing number of cases in which companies exaggerate their abilities to raise capital, according to Yann LeCun, the AI director at Facebook and a specialist in deep learning.