Artificial intelligence (AI) is about to be big business, according to the CSIRO's innovation hub Data61.\r\n\r\nThe research centre estimates that digital tech (including AI) will contribute an extra $315 billion in gross economic value to Australia over the next decade.\r\n\r\nThose projections were part of a sector roadmap released by Data61 on Friday, with direct support from the federal government.\r\n\r\nPolicymakers have outlined three core areas where AI and digital tech can add material value:\r\n\r\n- Health, aged care and disabilities; home-monitoring and data analysis to improve costs and efficiency of healthcare\r\n- Cities & infrastructure; automation and sensory systems to improve building infrastructure\r\n- Natural resources; mining site automation and agricultural robotics.\r\n\r\nRead: The move to automated mining is on, but who\u2019s really ready to join the robot revolution?\r\n\r\nWhile a government roadmap is one thing, the potential scale and relative nascence of the sector means putting together an investment roadmap isn't a simple task.\r\n\r\nThat being said, numbers don't lie. We commissioned the Stockhead data team to run some numbers on a group of 27 companies with exposure to AI as part of their business operations.\r\n\r\nOver the last 12 months, the cohort has made a robust average return of 29 per cent. Those gains were driven by 12 winners, while 17 finished in the red.\r\n\r\nLeading the pack was Resonance Health (ASX:RHT), which shot higher at this time last year after its AI product Ferrismart -- used to detect iron concentrations in the liver -- received approval from US regulators.\r\n\r\nBut perhaps as evidence that the sector is still finding its feet commercially, the average gain over the past six months fell to 14 per cent, with only eight companies in positive territory.\r\n\r\nHere\u2019s a table of ASX-listed companies with exposure to AI, and their 12-month share price performance.\r\n\r\nSwipe or scroll for full table. Click headings to sort\r\n\r\n\r\n\r\n \r\nInvestor view\r\nFor Dean Fergie, principal at small-cap fund manager Cyan Investments, assessing opportunities in AI still requires a degree of scepticism from a stock-picking point of view.\r\n\r\nSpeaking with Stockhead, Fergie said he'd noticed the increased use of AI by lending businesses, where companies are deploying the use of decision-making engines based on credit characteristics.\r\n\r\n"They'll compile plenty of spending data and the use of AI gets thrown around, but I think at this stage it's still difficult to tell how much it's a buzzword and how much it actually means," he said.\r\n\r\n"I wouldn't invest in a business that's using it purely as a basis for their whole technology proposition. But there's certainly businesses that have existing operations that can use an element of that to improve what they've already got."\r\n\r\nOne AI-related company in Cyan's portfolio is Alcidion (ASX:ALC), the data company who's core Miya Precision platform leverages AI-based predictive analytics in clinical support and patient management.\r\n\r\nAs a measure of how healthcare is seeing some early success among AI-based applications, Alcidion is the second best performer among AI-related ASX stocks with a 12-month gain of more than 300 per cent.\r\n\r\n"They utilise multiple inputs where a clinician would know what to do, but less experienced staff having smart software to help in the decision-making process can be really valuable," Fergie said.\r\n\r\n"But again, that's an existing operation where AI and machine learning forms part of the product offering. There's still a lot of historical data they're using.\r\n\r\n"I wouldn't want it to be the cornerstone of a company's value proposition if we haven't seen the evidence of commercialisation."\r\n\r\n \r\nCutting through the noise\r\nAnother company working on AI-based solutions in the health space is Coviu, an unlisted company that builds software which allows healthcare professionals to treat patients in remote locations via secure video linkup.\r\n\r\nCoviu CEO Dr Silvia Pfeiffer said that despite the potential for AI to foster improved health outcomes, developments in the technology should still be treated with a degree of caution.\r\n\r\nPfeiffer said getting clinical validation for the use of AI in healthcare "isn\u2019t yet common practice in the industry".\r\n\r\n"Too much AI is trained on biased data and this reinforces human preconceptions, including prejudices," he said.\r\n\r\n"When it comes to clinical decision support in particular, it\u2019s critical that the AI tools used by clinicians are unbiased."\r\n\r\nIn that context, she welcomed the continued efforts by the government to set out a regulatory roadmap for the sector, which she said would "better protect minorities in particular from faulty decisions of AI algorithms".\r\n\r\nOutside of healthcare, a primary catalyst for government interest in AI -- and information technology more broadly -- is that the sector is now a key employer in the Australian economy.\r\n\r\nThe CSIRO says more than 650,000 people hold IT jobs domestically, a number that's expected to grow by 20,000 per year through to 2023.\r\n\r\nBut for LiveTiles (ASX:LVT) CEO Karl Redenbach, a successful roadmap for continued growth in the sector will take more than just "basic aspirational guidelines".\r\n\r\nHe said strong policy in the area would require an accurate assessment of "what the future workplace could look like by harnessing the full potential of what AI technology can deliver".\r\n\r\n"Australia must realise the role it can play in shaping the conversation around the possibilities for AI, particularly in delivering improved end outcomes around productivity, new job creation and increasing business competitiveness," Redenbach said.