Glossary term
Glossary term
Governance and Compliance
Defined boundaries for how employees, contractors, and systems may use AI tools. Acceptable use guidance addresses sensitive data, prohibited tasks, human review, approved tools, and escalation for novel use cases. For generative AI, mature rules cover confidential data, client data, regulated data, code generation, external publishing, legal advice, and unapproved tools.
After Samsung engineers leaked source code and meeting notes via ChatGPT in March 2023, Samsung Electronics banned employee use of public generative AI tools on company devices.
Amazon's internal acceptable use guidance, leaked in early 2023, warned employees not to share confidential information with ChatGPT, later replaced by approved internal tools.
Anthropic's Usage Policies for Claude prohibit specific high-risk uses including generation of CSAM, weapons of mass destruction guidance, and content intended to manipulate elections.