Advances in generative AI come with many challenges for leaders. Forced to quickly take advantage of it while limiting its dangers, companies oscillate between fear and fascination.
To better understand how organizations are dealing with these new risks, BCG and the MIT Sloan Management Review looked at enterprise-wide Responsible Artificial Intelligence (IAR) programs . Thus, according to the results of this new report, the share of leaders who take an active role in the implementation of an IAR program has increased by 13 points in 2023, to reach 29% of the leaders surveyed.
However, how can these safeguards be adapted to the generative AI revolution? What are the objectives pursued by these ARI programs? With what results?
To answer this, BCG and the MIT Sloan Management Review surveyed more than 1,240 organizations in 59 industries and 87 countries. Among its main lessons:
- Increased maturity but challenged by the arrival of generative AI: the share of leading companies in IAR has almost doubled, rising from 16% to 29% between 2022 and 2023. Their share is higher in regulated sectors such as financial services, insurance, health and the public sector. Despite notable progress, 71% of organizations remain at the development stage or behind on these issues. The arrival of generative AI complicates the approach of companies. And for good reason, the testimonies collected point out that most ARI programs were not designed to cope with the sudden explosion of risks introduced by these new tools.
- A competitive advantage in sight: Leaders who take an active role in the implementation of an IAR program benefit from better returns from their IAR programs. Among them: better products and services (71%), increased innovation (66%) and better long-term profitability (57%).
- Dependence on external suppliers which raises questions: 78% of organizations use AI solutions developed by external suppliers. Even more: more than half (53%) rely exclusively on AI tools that were not designed or developed in-house. This use exposes them to a multitude of risks (financial, reputational, regulatory). However, these are only too rarely assessed (20% of organizations do not assess these risks at all).
The report thus identifies several ways to enable companies to equip themselves with a solid IAR program capable of adapting quickly to the challenges raised by generative AI. Among them, the involvement of the CEO, the systematic evaluation of third-party tools and the anticipation of new regulations.
Download report from here
