By Keith Strier and Cathy Cobey, EY

As the use of artificial intelligence (AI) and machine learning proliferates, AI technologies are rapidly outpacing the organizational governance and controls that guide their use.

External regulators simply can’t keep up, and enterprises are grappling with increasing demands to demonstrate sound and transparent controls that can evolve as quickly as the technology does.

As has been proven time and again in recent high-profile catastrophes, there are serious operational risks of using AI without a robust governance and ethical framework around it. Data technologies and systems can malfunction, be deliberately or accidentally corrupted and even adopt human biases. These failures have profound ramifications for security, decision-making and credibility, and may lead to costly litigation, reputational damage, customer revolt, reduced profitability and regulatory scrutiny.

The need to build trust

Within the organization, leaders must have confidence that their AI systems are functioning reliably and accurately, and they need to be able to trust the data being used. Yet this remains an area of concern; in our recent survey, nearly half (48%) of the respondents cited a lack of confidence in the quality and trustworthiness of data as a challenge for enterprise-wide AI programs.1

Meanwhile, organizations also need to build trust with their external stakeholders. For example, customers, suppliers and partners need to have confidence in the AI operating within the organization. They want to know when they are interacting with AI, what kind of data it is using, and for what purpose. And they want assurances that the AI system will not collect, retain or disclose their confidential information without their explicit and informed consent. Those who doubt the purpose, integrity and security of these technologies will be reluctant — and may ultimately refuse — to share the data on which tomorrow’s innovation relies.

Regulators are also looking for AI to have a net positive impact on society, and they have begun to develop enforcement mechanisms for human protections, freedoms and overall well-being.

Ultimately, to be accepted by users — both internally and externally — AI systems must be understandable, meaning their decision framework can be explained and validated. They must also be resolutely secure, even in the face of ever-evolving threats.

Amid these considerations, it is increasingly clear that failure to adopt governance and ethical standards that foster trust in AI will limit organizations’ ability to harness the full potential of these exciting technologies to fuel future growth.

You can get a copy of EY’s full report on the importance of trust in AI here.