Why Enterprise AI Risk Planning Must Look Beyond the Obvious

In a recent LinkedIn post, Clare Kitching discusses how companies assess the wider risks of artificial intelligence, arguing that business planning should look beyond the threats closest at hand. This article reports on Kitching’s post and analysis; it is not written by her.

“32% of the top 100 companies have not publicly acknowledged any risk from AI.”

Kitching cites a study by the MIT AI Risk Initiative, which, according to her summary, found that 32% of the top 100 companies had not publicly acknowledged an AI risk, rising to an estimated 62% among the top 1,000. She notes that companies’ public acknowledgements clustered around a handful of concerns: discrimination, weapons and cyberattacks, fraud and scams, privacy loss, and the devaluation of human creativity.

Risks can be harder to see when they feel distant

In Kitching’s view, organizations are more likely to take a risk seriously when they can picture its effects on their own operations. She points to power centralization, competitive dynamics, inequality, and unemployment as issues that may seem remote or uncomfortable, even though they could affect employees, customers, suppliers, and markets.

“We can take a risk more seriously when we can picture it happening in our own business.”

Kitching also highlights a gap between sector exposure and company recognition: she says 77% of risks went unacknowledged by companies in sectors experts judged “extremely vulnerable” to them. The post does not suggest that each company can resolve systemic challenges by itself. Instead, Kitching argues that businesses can include them in planning by identifying their exposure, considering how their decisions might worsen risks, and deciding what steps are available now.

Putting broader risks into business planning

As examples, Kitching raises dependence on a small number of AI providers, planning for changes to jobs, and setting limits on deployment speed when competitors are moving quickly. These examples connect abstract concerns to decisions companies already make about suppliers, staffing, and technology adoption.

Taking risk seriously without alarmism

Kitching opens by recalling a wave of LinkedIn posts urging people to stock up on tinned beans and water because of AI risks. She calls that reaction excessive, while making a separate case for measured enterprise preparation. Her closing line captures the distinction:

“You don’t need to stock up on tinned beans to take AI risk seriously. But you need to look beyond the risks closest to your desk.”

The post’s central point is that responsible AI risk planning need not depend on panic. According to Kitching, it starts with widening the frame: alongside familiar, immediate concerns, companies should examine less visible risks and consider how those could shape the businesses and communities around them.