The cloud-based service will detect bias and also explain how the artificial intelligence (AI) made the decision, IBM Research Vice President India and Singapore and CTO IBM India/South Asia Sriram Raghavan said.
In addition, IBM Research will also release into the open source community an AI bias detection and mitigation toolkit that will help developers design features to address bias in AI.
Raghavan said two of the biggest barriers in adoption of AI was lack of trust and concerns around compliance.
"With this service, we want to address both of these. It will bring in transparency as customers will know why the AI system took a certain decision. It will also detect bias that may come into decisions on account of multiple reasons," he said.
IBM is giving new transparency and control to the businesses who use AI and face potential risks from any flawed decision making, he added.
Raghavan said one of the ways that bias can creep in is with data input into the AI system.
"For example, if a financial institute historically has avoided lending to, say those under 25 years, the AI system may process it as a rule and while making a decision, take that into account," he explained.
The fully automated software service explains decision-making and detects bias in AI models at run time – as decisions are being made – capturing potentially unfair outcomes as they occur.
It also automatically recommends data to add to the model to help mitigate any bias it has detected.
All of these capabilities are accessed through visual dashboards, giving business users an unparalleled ability to understand, explain and manage AI-led decisions, and reducing dependency on specialised AI skills.
