I recently read a UC Berkeley article that put words to something I believe about business: Focus on profit, but never lose sight of the responsibility to do right.
That responsibility belongs in the decisions we make about artificial intelligence.
Speaking at Berkeley, Nobel Peace Prize laureate Kailash Satyarthi urged students to embed compassion in AI. He described compassion as more than kindness or empathy. It includes “mindful, pragmatic problem solving.”
That caught my attention. Compassion sounds abstract until we ask what it requires of the people building, buying, and using the technology.
Who benefits? Who could be harmed? Who takes responsibility?
For the smaller employers I serve, those questions belong alongside the discussion about saving time and reducing costs.
What Are We Trying to Improve?
Consider a small employer evaluating AI to answer employee benefit questions.
I would want that employer to define success before choosing the tool. Is the objective to answer more questions, reduce administrative work, help employees understand their benefits, or some combination?
Then I would ask what happens when the tool gives an incorrect answer.
Can an employee reach a person? Does someone review the information? Is sensitive employee information protected?
My view is that a useful tool should be evaluated on both the work it completes and the consequences of its mistakes.
That is where compassion becomes a management decision.
The Connection to ESG
I see a direct connection to environmental, social, and governance considerations.
The social questions concern people: privacy, fair treatment, access, and the effects on employees and customers.
The governance questions concern responsibility: Who approves the technology? Who checks its work? Who can stop its use when something goes wrong?
The environmental questions concern the resources used to build and operate the technology, and whether those impacts are being assessed.
These concerns have a practical foundation. The National Institute of Standards and Technology’s AI framework addresses privacy, harmful bias, accountability, transparency, and environmental impacts.
For investors, I would frame the question this way: How well does a company understand and manage the consequences of the technology it uses?
Calling a system “responsible” tells me little. I want to understand the practices behind the description.
Doing Right When There Is a Tradeoff
It is easy to support profit and compassion when they point toward the same decision.
The harder question is what happens when protecting people requires more time, more oversight, or a decision to forgo a particular use.
I believe doing right should shape how a company earns its profit. It should influence the choices made before a problem occurs.
For a smaller employer, I would start with five questions:
- What specific problem are we trying to solve?
- Whose information will the tool receive?
- Who could be affected by an incorrect answer?
- Which decisions require human review?
- How will we determine whether the tool is helping?
The Berkeley article offers an ethical argument, rather than evidence that compassionate AI produces higher investment returns. Its value is in challenging us to examine the objectives we give technology and the standards we use to judge success.
The Parting Glass
Profit matters. So does the way we earn it.
My standard for adopting AI would include the time saved, the quality of the work, and the treatment of the people affected.
Compassion belongs in those decisions from the beginning.
#ResponsibleAI #ESGInvesting #SmallBusiness #CorporateGovernance #EmployeeBenefits