This role represents a new job category that emerged with large language models: specialized enforcement analysts who review AI model interactions in real-time to detect and prevent specific categories of misuse (in this case, biological harms). This combination of biosecurity threat analysis with AI platform enforcement didn't exist before LLMs became powerful enough to pose dual-use risks.

This role represents a new job category focused specifically on managing the intersection of frontier AI capabilities and biosecurity/biotech policy—a concern that emerged only as large language models demonstrated potential biological knowledge risks, making this a distinctly post-2022 policy specialization.

This role represents a new job category focused specifically on enforcing safeguards against AI-enabled harms like coordinated inauthentic behavior, election manipulation, and AI-facilitated surveillance—work that didn't exist before large language models became widely deployed.

Red teaming for AI safety—systematically probing advanced AI systems for novel vulnerabilities and emergent risks unique to AI capabilities—is a fundamentally new job category that didn't exist before large language models became deployed products.

Harvey is hiring its founding customer success lead dedicated to law schools — a legal-AI vendor building a distinct sales motion for legal education. As the first hire in that segment, this person defines how the product embeds into curricula, clinics, and research.

This role combines building and maintaining demonstration infrastructure for frontier AI capabilities with technical storytelling—a specialized function that emerged specifically to bridge the gap between cutting-edge AI products and customer understanding, which didn't exist as a distinct job category before the recent wave of generative AI products.

This role represents an emerging category of AI-native operations specialists who design and deploy AI-powered workflows to transform traditional business processes. It combines revenue operations with frontier AI product development and workflow engineering—a job category that has only emerged as organizations begin systematically applying large language models and AI agents to operational workflows.

Managing threat intelligence specifically for AI model exploitation, distillation networks, and fraud against large language models is a new job category that emerged with the proliferation of commercial AI systems—this specialized investigative work targeting AI-specific attack vectors didn't exist before widespread LLM deployment.

This role represents a new job category: managing threat intelligence specifically for AI system misuse detection, requiring expertise in CBRN-E and advanced weapons proliferation in the context of generative AI safety—a responsibility that didn't exist before large language models became accessible tools.

This role represents a new job category focused on detecting and disrupting misuse of frontier AI systems for influence operations and surveillance—work that didn't exist before large language models became accessible tools for potential abuse by state and commercial actors.

This role focuses on systematically shaping AI model personalities and behaviors through reward modeling and reinforcement learning—a specialized research discipline that emerged specifically with large language models and didn't exist as a distinct job category before 2022.

This role represents a new job category focused specifically on building safeguards and risk mitigation systems for generative AI deployment—work that didn't exist before large language models became powerful enough to require dedicated safety infrastructure and governance.

This role represents a new category of 'AI Acceleration Engineer' focused on systematically integrating frontier AI tools and agents into company operations and development processes—a discipline that has emerged only in the last 2-3 years as AI capabilities have become production-ready.

This role focuses on security and privacy threats specific to frontier AI systems and AI-native products (autonomous agents, LLM-based search), a threat landscape and defensive research domain that didn't exist before the recent emergence of large language models and generative AI systems.