Many are eager to harness the increasingly sophisticated capabilities of ChatGPT and generative artificial intelligence (AI).
According to Baker McKenzie’s 2022 North America AI survey, business leaders may not be fully aware of the risks associated with artificial intelligence. The number of C-suite leaders who are concerned about the risks associated with AI is only 4%, and less than half of them have board level AI expertise.
The figures highlight a troubling reality: many organizations lack the oversight and expertise needed to manage risks associated with AI. Organizational blind spots surrounding the technology’s ethical and effective deployment are likely to overshadow transformative opportunities and cause organizations to fall behind the technology’s explosive growth if left unaddressed.
What are the implications of generative AI for the risk landscape?
Currently, AI-related progress and adoption are accelerating exponentially – some argue too quickly.
In recent years, academics, scientists, policy-makers, legal professionals, and others have been advocating for AI’s ethical and legal applications. AI is already being used to recruit talent, perform administrative duties, and train employees at the workplace.Seventy-five percent of companies already use AI tools and technology for hiring and HR.
Governance, accountability, and transparency are more important than ever in this new phase of generative AI, as are concerns over poorly deployed AI’s consequences.
Data privacy and breaches are another concern, easily occurring through the non-anonymization and collection of employee data. Unchecked algorithms can result in biased and discriminatory outcomes, perpetuating inequity as well as dampening progress on workforce diversity.
The advent of generative AI has also provided new IP considerations, raising questions about ownership of inputs and outputs from third-party programs, and subsequent concerns about copyright infringement.
AI-related legislation and regulatory enforcement mechanisms have been scrambling to be adopted by governments and regulators. In the US, emerging legislation will focus on using AI for recruiting and HR-related operations.
We have already seen the first wave of generative AI IP litigation in the US, and these early court decisions are reshaping the legal landscape in the absence of existing regulations.
Data fed into AI tools and queries may also be collected by third-party providers of the technology. In some cases, these providers have the right to use and/or disclose these inputs.
Are employers putting sensitive data and trade secrets at risk as they integrate generative AI tools into their workforces? In short, yes. Every new development seems to raise more questions than organizations, regulators, or courts can answer.
What can organizations do to improve their AI readiness?
To stay ahead, organizations will need to move beyond siloed efforts and bring together discrete functions under the umbrella of a strong governance framework. Generative AI is changing the paradigm, and specific use cases will continue to raise risks.
All relevant stakeholders, including legal, the C-suite, boards, privacy, compliance, and HR, need to be involved throughout the decision-making process, not just data scientists.
According to our survey findings, only 54% of respondents said that their organization involves HR in the decision-making process for AI tools, and only 36% said that their organization has a Chief AI Officer (CAIO).
HR must be involved in training and fostering a cross-functional AI team in this high-risk environment, as the CAIO will play a critical role in ensuring appropriate governance and oversight at the C-Suite level.
Moreover, organizations should develop and follow an internal governance framework that takes enterprise risks into account across use cases and allows them to efficiently address compliance issues.
Companies without an AI governance structure or using third-party tools exclusively run the risk of utilizing AI tools in a way that creates organizational legal liability (e.g., discrimination claims).
A bias exists in virtually all decision-making, whether it is artificial intelligence or not. In order to meet data privacy requirements, companies that use these tools must develop a framework that identifies an approach to assessing bias as well as a mechanism to test and avoid unlawful bias.
Effective measures for pre- and post-deployment testing should further support efforts to combat bias.
AI deployment companies must also ensure that they have processes in place that provide an understanding of the data sets being used, algorithmic functionality, and technological limitations, since proposed legislation will likely require reporting.
The Final Outlook
AI Offers many advantages. However, its rapid implementation and growth have elevated the importance of strategic supervision and governance for responsible usage and risk management.
The tendency to adopt AI without proper precautions is alarming, as many organizations are unprepared and underestimate its potential hazards.
Nevertheless, by establishing robust governance and oversight mechanisms, organizations can navigate the technological advancements at any stage of their AI journey.
As a result of the risks associated with artificial intelligence, well-informed individuals from the legal, regulatory, and private sectors will have to collaborate in the future. As a result, legislation, codes of conduct, and guidelines can be developed that acknowledge both the benefits and risks of this technology.
Organizations can deploy AI technology more confidently with a secure framework in place.