The business world is barreling into the era of Generative AI, a technology poised to unlock higher human potential, making them more knowledgeable, creative, and productive. Yet there’s another aspect to consider. GenAI can potentially create new data privacy and ethical use case risks, meaning IT leaders should establish guardrails to ensure responsible use.
Organizations—and individuals—are embracing GenAI in earnest. A 2024 McKinsey Global Survey on AI found 65% of respondents reporting regular use of GenAI in their organization, nearly double the percentage from a survey conducted 10 months prior. Expectations for the technology remain high—three quarters of McKinsey survey respondents anticipate GenAI will lead to disruptive change in their industries in the years ahead.
As GenAI adoption soars, business and IT leaders can’t assume the ethical ramifications of the technology are widely understood. CIOs, CDOs, and CISOs need to partner with their HR and legal counterparts to define clear policies, communicate expectations, and establish formal AI governance practices.
“Leaders can’t expect that employees know what constitutes ethical and responsible use of GenAI in their organizations in the context of regulations, the data they collect, customer expectations, and company culture,” says Issac Sacolick (@nyike), president of StarCIO and author of Digital Trailblazer. “Responsible GenAI requires organizations to define clear business objectives, classify data, secure access to data, establish which GenAI tools can be used, and quantify quality and accuracy metrics.”
Gene de Libero, principal at marketing technology consultancy Digital Mindshare LLC, advises companies to get started with the basics. This includes creating rules around how AI is used and mechanisms that check GenAI output to ensure they are fair and unbiased. Systems can play a role, addressing biases through data input refinement and diverse contextual datasets, but importantly, they are reinforced by human input, adds Arsalan Khan (@ArsalanAKhan), a speaker, advisor, and blogger.
GenAI has an insatiable appetite for data, but organizations must be mindful about what data is included and the long-term ramifications of its use. Misuse of sensitive data without explicit consent is one of the biggest concerns. For example, doctors can greatly benefit from GenAI use in healthcare scenarios, but a patient’s condition and prescription history can live forever in a GenAI model, notes Peter Nichol, data and analytics leader for North America at Nestlé Health Science. Similarly, financial advisors sharing specific financial details with GenAI tools to garner better insights must consider privacy laws such as General Data Protection Regulation (GDPR) or the California Consumer Privacy Act (CCPA) – and others, depending on the operating country impacted, Nichol says.
A roadmap for success
Before GenAI tools become too entrenched, organizations should double down on these strategies to establish guardrails and ensure responsible and ethical AI usage:
Raise awareness about how the technology works. Most have heard about GenAI and its promise to help people work smarter, but only a fraction of potential users know the particulars of how large language models (LLMs) work. What most don’t understand is that any private or confidential data submitted to a publicly available LLM like ChatGPT is at risk for public exposure, explains Chris Selland, partner at Tech CXO. “For the same reason, developing and provisioning a private GenAI platform encompassing robust security measures is crucial,” he says.
Establish clear communications. Transparency becomes more important as GenAI becomes a fixture in organizations. Companies must clearly communicate how and where they use AI, especially when it affects customer interactions or is the foundational mechanism for creating content. “Being upfront about AI’s role builds trust and helps customers understand its impact on their experience,” says Scott Schober (@ScottBVS), president and CEO at Berkeley Varitronics Systems Inc.
Communicating clearly about ethical guidelines within the internal organization is equally crucial. “Organizations should proactively develop policies that define acceptable uses of GenAI, rooted in what’s generally accepted in society,” Schober adds. “This ensures responsible innovation while maintaining alignment with the company’s culture.”
Create an AI council or governance body. Establishing a formal body that takes responsibility for AI technology developments and governance practices is important. The group should be tasked with ensuring any resulting AI applications meet corporate ethical standards and regulatory requirements, explains Ben Rothke (@benrothke), senior information security manager at Tapad.
In addition, AI councils should include representation from different parts of the business as well as diverse voices from within and outside the organization. The goal is to have a constituency that challenges core assumptions and ensures broader societal alignment.
Adopt a flexibly strategy. As a whole, AI safeguards must be resilient, meaning they have staying power to flag unethical content even as the technology and use cases evolve. Ethical guardrails should be embedded directly into GenAI workflows to ensure nothing operates outside of predefined constraints. Finally, maintaining an AI audit trail is important—these practices and systems ensure every AI decision is transparent, explainable, and revisitable as opposed to operating as an AI black box.
The bottom line
Given the hyper speed at which GenAI is unfolding, it’s important to recognize that ethical considerations are not static. “It requires iterative feedback loops to refine behavior as societal norms evolve,” cautions Shail Khiyara, CEO at Plutoshift AI and founder of the VOCAL Council.