Nearly half of the companies in Europe that call themselves AI start-ups don't in fact use artificial intelligence, a new report found.
The research, published Tuesday by London-based venture capital firm MMC Ventures, found no evidence that artificial intelligence was an important part of the products offered by 40 percent of Europe's 2,830 AI start-ups. The report's authors individually reviewed the activities, functions and funding of start-ups across 13 EU countries. It did not name any of the start-ups involved in the study.
The findings raise questions about how the term AI has become a blanket phrase for start-ups looking to attract investments and position themselves at the forefront of tech innovation.
"There's a huge amount of hype around, but one of the ways that we hope to add value as an investor … is to help cut through that and highlight the technology," said David Kelnar, head of research at MMC, in a phone interview with CNBC on Tuesday.
Kelnar said Europe is entering an "era of AI entrepreneurship" with one out of every 12 new start-ups putting AI at the core of their operations. In 2013, the survey found just one in 50 new start-ups embraced AI.
One reason start-ups could be calling themselves AI companies is to attract more funding. MMC's survey found AI companies have raised between 15 and 50 percent more capital in their funding rounds than traditional software companies since 2015. AI start-ups also secured higher valuations than software firms.
Peter Finnie, a partner at intellectual property law firm Gill Jennings and Every, said there has been a surge in AI patent filings as companies look to make their businesses more investable.
"From an intellectual property perspective, the use of AI in this way actually holds little to no value," he said in an email to CNBC. "The AI sector is a bubble that might well burst because of this."
Companies and regulators say one problem is the lack of a unified definition of artificial intelligence. The European Commission, the EU's executive arm, broadly defines AI as a technology that allows systems to analyze their environment and take decisions "with some degree of autonomy." The Commission is currently working on a set of guidelines around AI to help establish clear definitions and ensure it is used for "ethical" purposes.
"I think it will actually help us to be more competitive than others are, because users will trust AI made in Europe versus AI made elsewhere," the European Commission's Director of Artificial Intelligence Lucilla Sioli told CNBC's Karen Tso last week. "It is very important that we get this approach right."
Within Europe, the United Kingdom is home to the most AI start-ups, according to MMC's report. The U.K.'s profile has been raised by start-ups like DeepMind, a British research firm acquired by Google in 2014.
Health care is the most popular sector for AI start-ups in Europe, with one in five new companies focusing on health and wellbeing. Finance, media and entertainment and retail were also popular industries for AI, according to the MMC report.