AI is making the big leap from laboratory to real world – to much fanfare and hype around either its benefits, or the ravages it will supposedly wreak on our lives. To gain buy-in from staff and customers, business will have to complement AI technology with governance, ethics and trust. Can we be certain AI is sticking to the rules –and has it even been informed what those rules are? Here’s how companies can make a start with AI ethics:
- Begin with the data
Make sure it’s accurate and free of bias, and establish checks and balances to ensure data is being used in compliance with regulations (such as GDPR in the EU). It’s also important to ensure AI is learning from the right inputs to keep its decisions smart and valuable.
- Make the logic transparent
Black-box-style decisions by AI systems are unlikely to inspire trust, and problems easily arise if the logic is unclear. Credit-scoring systems, for instance, can pick up gender and race bias built into data – and perpetuate this bias in their decisions.
- Train AI to behave ethically
What are companies doing to prevent AI behaving badly? According to a Cognizant survey, around half have no official policy in place. Yet as AI gathers pace, business will need to reinforce efforts to keep AI systems in lockstep with accepted practices for good governance. At a simple level, this could mean no passing on responsibility to robots: the buck always must stop at a human desk.
- Re-educate the workforce
AI will spawn completely new jobs (check out our “21 Jobs of the Future” report for ideas) that rely on human creativity and judgement. Other jobs, however, like front-line tech support, telemarketers and legal assistants, are under threat. Business will play a key role in promoting life-long learning, and encouraging people to acquire new skills for an AI-powered economy.
Rolling out limited-scope AI applications is relatively straightforward. The real challenge comes when AI moves from the edge of the business to its core. To build trust that drives success, companies will have to keep the big issues of AI ethics and governance in mind.
To learn more about the key ingredients of successful AI – real-life business cases, and the role of good governance – download the Cognizant report Making AI Responsible – And Eﬀective.