Leaders should steer for responsible use of ChatGPT
According to Gartner, legal and compliance leaders should address the following six risks associated with the use of ChatGPT and other large language model (LLM) tools:
- Fabricated and Inaccurate Answers: ChatGPT may provide incorrect information that appears plausible. Organizations should establish guidance requiring employees to review the output generated by ChatGPT for accuracy before accepting it.
- Data Privacy and Confidentiality: Information entered into ChatGPT, if chat history is enabled, may become part of the training dataset and be accessible to users outside the enterprise. Legal and compliance leaders should establish a compliance framework and prohibit the entry of sensitive organizational or personal data into public LLM tools.
- Model and Output Bias: Despite efforts to minimize bias, ChatGPT may still exhibit biases. Legal and compliance leaders need to stay informed about laws governing AI bias and ensure compliance by working with subject matter experts and implementing data quality controls.
- Intellectual Property (IP) and Copyright Risks: ChatGPT is trained on internet data that may include copyrighted material, posing potential violations of copyright or IP protections. Organizations should monitor copyright law changes and require users to scrutinize output to avoid infringement.
- Cyber Fraud Risks: ChatGPT can be misused by bad actors to generate false information or perform malicious tasks. Legal and compliance leaders should coordinate with cybersecurity personnel and conduct due diligence audits to verify the quality of information.
- Consumer Protection Risks: Organizations that use ChatGPT, such as customer support chatbots, must comply with relevant regulations and disclose clearly and conspicuously that a consumer is interacting with a bot. Failure to do so may result in loss of customer trust and legal consequences.
Legal and compliance leaders should assess these risks and establish appropriate guardrails to ensure responsible enterprise use of generative AI tools. Failure to address these risks may expose organizations to legal, reputational, and financial consequences.
