Report: 66% of AI decision-makers are concerned about meeting ethical business goals

According to a new report by InRule Technology, based on data from a commissioned survey with Forrester, 66% of AI/ML decision-makers reported concerns about an inability to meet their organization’s ethical business goals. Other top risks cited include high costs (60%), bad business outcomes (47%) and failure to meet regulatory requirements (37%). Only slightly more than half (56%) are confident in their ability to mitigate these risks.

Artificial intelligence (AI) and machine learning (ML) are critical in the success of digital decision-making, according to nearly two-thirds (66%) of decision-makers, with this number expected to jump to 95% in three years. Still, the complexities of adopting a range of technologies for automated decisioning processes is causing roadblocks.

Robotic Process Automation (RPA) and Digital Process Automation (DPA) are common, but not fully meeting the standards of “automatic”: 70% of users cited difficult maintaining complex decisions using DPA-only solutions and 41% said their organization uses a business process management system with an external decision engine. In other words, enterprises are clamoring for automation solutions, but are getting stuck dealing with disparate technologies that don’t always integrate seamlessly and require constant evaluation and oversight. This is evident in nearly two-thirds of respondents, reporting that technological, organizational and operational problems impede agility and scalability, especially scope creep and model drift.

While IT decision-makers understand the value of AI/ML in decision-making technologies, respondents said they also worry that harmful bias can lead to inaccurate decisions (58%), inconsistent decisions (46%), decreased operational efficiency (39%) and loss of business (32%). As IT decision-makers express fear of damaging customer relationships, it’s not an unfounded fear: more than three-fourths also said their organization has already experienced harmful bias in its digital decisioning initiatives.

Read the full report by InRule Technology.