Insurance-industry actuaries are highly specialised at quantifying risk – but as an insurance consultancy snaps up an artificial intelligence (AI) innovator, actuarial algorithms will soon be evaluating everything from your medical risk factors to your susceptibility to online advertising.

Actuarial and analytics consultancy Finity Consulting bought long-term partner Deep Logic – a Sydney-based AI firm that provides its Neuronworks-branded solutions to top-20 insurers across the US, UK, and Europe – in what it sees as a natural fit for a business that is riding surging data trends to look beyond its core insurance customer base.

AI techniques will allow insurers and other businesses to fine-tune their risk calculations to accommodate changing market dynamics.

Yet as AI-powered actuarial capabilities expand risk management outside of its traditional insurance base, the field of data science may need to follow global guidance by addressing their ethical and moral implications.

Social advocates worry that growing data stores, and the increasing specificity of AI algorithms, are helping businesses justify discrimination against higher-risk groups.

Availability of “reasonable” actuarial or statistical is a key factor in legitimising legally acceptable discrimination, the Australian Law Reform Commission has noted.

Many Australians have had life insurance denied or prejudiced by genetic testing, and mental health support groups have been lobbying to stop insurance providers from discriminating against Australians with mental health issues.

Industry group the Financial Services Council has long offered guidelines for insurers’ use of genetic testing data – and radically changed them this month by issuing a five-year moratorium on the use of genetic tests in processing applications for less than $500,000 of life insurance.

Staking a claim in AI

However its technology is ultimately used, Finity’s acquisition of Deep Logic is a proactive step to expand into new industries – and to secure highly specialised analytics and AI skills in a market where they’re increasingly hard to find.

“Technology companies have been the dominant deal makers in the AI space,” GlobalData financial deals analyst Aurojyoti Bose noted in the wake of a new analysis that found Facebook, Microsoft, Apple and Splunk together acquired 30 AI specialists between 2014 and 2018, while Accenture alone bought six others.

“However, with AI making inroads into diverse sectors, the buyer universe is expanding and the space is also attracting investments from non-technology companies.... Corporates are extensively evaluating options to integrate AI in their business operations.”

Deep Logic’s three staff – Dr Paul Beinat, Eddie Owen, and Nikolay Nikolaev – have joined Finity as partners, giving them an organisational gravitas that managing director Scott Collings said reflects the urgency with which AI is now perceived.

“It is hard to hire these skills, and really hard to know what you’re getting when you do,” Collings told Information Age.

“There is no well-defined qualification as to what a data scientist does, which makes recruiting quite difficult. But I think having the people who can do the smartest work and direct others, is the most important thing.”

Exploding interest in AI systems is driving enthusiastic business adoption of the technology, with IDC recently forecasting that worldwide spending on AI systems will reach $US35.8b ($A51.3b) this year – up 44 per cent on last year’s figures.

That adoption will grow an average of 38 per cent annually through 2022 – presaging a skills crunch that is already posing supply-side challenges given that data scientists are already topping recruiters’ lists of the most in-demand IT roles.

Bringing actuarial expertise outside of insurance

Around a third of Finity’s 150 employees already focus on data analytics and statistical modelling, so adding three staff is hardly going to make a ripple organisationally.

Yet the expertise the team brings – Beinat, for one, is an adjunct professor with the UTS Advanced Analytics Institute and linked with its Centre for Artificial Intelligence – will expand Finity’s risk-assessment capabilities into new industries such as healthcare and marketing.

“Understanding how to differentiate between two different types of people, and what sort of risks they represent, has always been a challenge,” Collings said.

“Ten years ago, you would have said that actuarial techniques were only applicable to the insurance industry but going forward it’s becoming a lot more blurry in terms of where those skills can be applied.”

Collection of massive volumes of operational and customer data had given firms in other industries a similar level of data focus, and AI would increasingly be sought as a way of turning this into business insight.

A recent PwC survey of more than 1,000 executives found that 20 per cent reported plans to adopt AI enterprise-wide this year – and data would be the fuel for the fire.

“There’s a whole smorgasbord of data for businesses to look at and decide how to use,” he explained, “and the availability of data, plus the ability to leverage it, has become a source of competitive advantage.”

“But you can’t trawl through these enormous databases easily; you need sophisticated machine-learning [ML] methodologies to make sense of it and help make what are increasingly live decisions with that data during a customer contact.”

Turning data into insight

Yet even in businesses with some analytics capabilities, turning data into timely decision-making can be a challenge: a recent Ivanti survey of over 400 IT professionals, for one, found that 51 per cent said they have to massage data for days, weeks, or more before it is actionable.

Deep Logic’s latest tool, called MySelf, aims to address this disconnect with an ‘artificial immune system’ designed to detect anomalies using a self-learning engine, which has already been tested on non-insurance applications such as digital ad fraud and cybersecurity.

“These are applications that have nothing to do with the kinds of industries we typically work in,” Collings explained.

“But they’re a perfect opportunity to apply AI concepts at the very highest level – and the ones who can leverage that data better will come out in front.”

Having secured a strong AI capability, Collings believes Finity is well-positioned to build and maintain a lead in applying AI to other risk-assessment processes – and anticipates bringing more staff on board to support their efforts.

“Insurers have always pushed the envelope because it’s a competitive advantage,” he said, “and Deep Logic are at the pointy end of expertise in an industry that’s pretty mature in its use of those skills already.”

“The important point is that we actually know what to do with them.”