C-suite executives are overconfident about battling AI threats at a time when U.S. regulatory and enforcement agencies are increasingly scrutinizing the harmful effects of AI, according to a survey.

The report from Baker McKenzie said only 4% of respondents thought AI risks were “significant.” But if something goes wrong, more than half of the respondents cannot override outcomes resulting from the use of AI. Moreover, a quarter of respondents say their company’s policies for handling AI risks are not documented or non-existent.

Around 36% of respondents did not have a CAIO in place. The ones who did not have a CAIO were also not in a hurry to hire one.

At the board level, around 38% believe their directors do not fully understand how AI is implemented across the company. Only 41% of those surveyed said their board members had expertise in AI.

Notably, 54% of those surveyed involve HR in the decision-making process for AI tools. However, the survey found critical weaknesses in oversight in HR and hiring tools.

Going forward

In the future, Baker McKenzie recommends carrying out a risk assessment of AI tools to manage the regulatory compliance, use, audit, and reporting requirements of the tools. Companies should also develop AI oversight and governance policies to reduce the risks AI can cause. Organizations should make sure AI tools are fair, transparent, and explainable while also respecting data privacy laws and ensuring accountability.

Finally, companies should tap experts familiar with the mindset of law enforcement and regulators in the U.S., U.K., Europe, China and other jurisdictions to ensure compliance with coming new regulations.

The study surveyed 500 C-level executives in the U.S. Baker McKenzie collaborated with Coleman Parkes to carry out the survey.