This month, G2 added five new categories to the taxonomy, and the throughline is hard to miss: Software is quietly graduating from helping people work to doing the work itself. Agentic GTM Platforms are taking over prospecting and pipeline orchestration. AI Legal Operating Systems are executing multi-step legal workflows, not just drafting documents. A whole new infrastructure layer — MCP Server Infrastructure Platforms — has emerged just to keep those agents safely wired into the rest of the enterprise stack.
Meanwhile, a fast-growing OpenAI implementation ecosystem is helping companies actually operationalize all of this, and Cyber Risk Quantification is giving security teams a way to translate the risk of getting it wrong into dollars the boardroom understands.
Five categories, one pattern: The tools aren't just getting smarter, they're getting agentic. Here's what's new on G2 this month.
About this category: Agentic GTM Platforms use AI agents to coordinate, execute, and optimize work across the go-to-market lifecycle, from identifying target accounts and researching prospects to orchestrating engagement, managing pipeline, and expanding customer relationships.
Why it was introduced: “AI is changing the way companies identify, engage, convert, and retain customers. As organizations look to accelerate growth while operating more efficiently, go-to-market technology is evolving beyond tools that simply assist sales and marketing teams toward platforms capable of actively carrying out work on their behalf.
A new class of Agentic GTM Platforms is emerging around this shift. Rather than adding AI assistants or generative features to existing sales and marketing workflows, these platforms use agents to interpret customer and market signals, make decisions, initiate actions, and coordinate multi-step workflows across prospecting, account research, ICP development, campaign execution, customer engagement, pipeline management, and revenue expansion.
G2 created this category because traditional sales, marketing, revenue intelligence, and customer success categories do not adequately capture platforms designed to operate across these functional boundaries. Agentic GTM Platforms increasingly act as an intelligent execution layer across the broader revenue organization rather than serving a single team or point solution.
This distinction is becoming increasingly important as GTM organizations move away from collections of disconnected tools and toward interconnected systems where AI agents can work across customer data, business context, applications, and channels. The market is beginning to separate vendors that layer AI capabilities onto existing software from those whose products are designed around autonomous, goal-oriented execution from the ground up.
For CROs, CMOs, revenue operations leaders, sales leaders, and growth-stage founders, the buyer need is also changing. These teams are looking beyond incremental productivity gains toward systems that can help determine who to target, prioritize opportunities, orchestrate engagement, respond to changing signals, and execute revenue workflows more efficiently.
Agentic GTM Platforms gives buyers a dedicated place to evaluate this emerging class of technology while creating a clearer home for vendors that span traditional sales and marketing categories but collectively represent a new operating model for go-to-market execution.”
- Bijou Barry, AI Principal Analyst
The top three products in this category, ranked by popularity at the time of launch, are:
About this category: AI Legal Operating Systems are agentic platforms that coordinate and execute legal work across matters, knowledge, applications, data sources, and stakeholders while maintaining the legal context, controls, and oversight required for enterprise use.
Why it was introduced: “AI adoption within legal teams is moving beyond individual productivity tools. While the first wave of legal AI focused heavily on isolated tasks such as drafting, research, summarization, and document review, organizations increasingly need systems capable of coordinating work across the broader legal environment.
AI Legal Operating Systems represent this next stage of the market. These platforms can understand legal and business context, determine appropriate next steps, execute multi-step workflows, interact with connected systems, and maintain context as work progresses. Rather than assisting with a single task, they increasingly function as an execution layer across legal operations.
G2 created this category because existing legal AI assistants and traditional legal software categories do not adequately distinguish this emerging architecture. Legal practice management and enterprise legal management platforms primarily serve as systems of record for matters, documents, billing, and operations, while AI Legal Operating Systems are designed to actively initiate, coordinate, and carry out work.
The distinction from horizontal AI agent platforms is equally important. Legal work requires specialized context surrounding matters, privilege, conflicts, confidentiality, approvals, regulatory obligations, and other legal-specific constraints. Platforms operating in this environment must combine agentic execution with human oversight, permissioning, auditability, and access to trusted legal knowledge and systems.
As legal departments move from experimentation toward broader AI adoption, the ability to connect systems and automate multi-step work will become increasingly important. The emerging opportunity is not simply to make lawyers faster at individual tasks, but to redesign how legal work moves across people, information, and applications.
AI Legal Operating Systems gives buyers a dedicated place to identify platforms built for end-to-end, agent-driven legal execution and helps distinguish this emerging market from standalone AI assistance and traditional legal systems of record.”
- Bijou Barry, AI Principal Analyst
The top three products in this category, ranked by popularity at the time of launch, are:
About this category: MCP server infrastructure covers the runtime, registry, gateway, and security tools that build, host, discover, and govern Model Context Protocol (MCP) servers. MCP is an open standard, originated by Anthropic and now stewarded by the Linux Foundation, that lets large language models (LLM) and AI agents connect to external data sources, tools, and services through a uniform interface. These platforms sit between AI agents and the backend systems that those agents act on. They handle authentication, tool discovery, audit logging, and policy enforcement for every tool call a model makes.
Why it was introduced: "MCP launched in November 2024. In 18 months, it went from one vendor's open-source spec to the layer AI agents now use to reach the rest of the enterprise stack. Hundreds of public MCP servers exist, the Linux Foundation hosts an official registry, and OpenAI, Google, Microsoft, and most major developer tools have adopted the standard. Buyers are asking which servers are safe to run in production, who hosts them, and how to govern what an LLM can call. G2 has no surface built to answer those questions yet. Buyers default to GitHub stars and community directories.
This category covers infrastructure, not the protocol itself. Any product can bolt on an MCP endpoint the same way any product can expose a REST API, and that alone doesn't qualify it. MCP Server Infrastructure is reserved for registries, hosting platforms, gateways, SDK generators, and security tools built specifically because MCP exists. Their primary job is running the lifecycle of an MCP server: connection negotiation, authentication, schema discovery, argument validation, execution, and audit.
Buying this category means coordinating three teams, the same complexity G2 already sees in API Management and Observability. AI engineering teams need to expose internal systems to agents without hand-rolling auth for every tool. Platform engineers own the uptime of anything an agent now depends on in production. Security and compliance teams need an audit trail for what an LLM called, when, and with what data, before agents touch production systems. As agentic AI moves from pilot to production, this category becomes the control layer that decides whether that move is safe."
— Sohan Pal, Research Analyst
The top three products in this category, ranked by popularity at the time of launch, are:
About this category: OpenAI Consulting Services are firms that help enterprises design, build, deploy, and scale solutions specifically using OpenAI’s models, products, and development platforms.
Why it was introduced: Enterprise adoption of OpenAI is moving beyond experimentation and isolated use cases. Organizations are increasingly deploying OpenAI technologies into production environments where implementation, integration, governance, security, workflow design, and organizational adoption become as important as access to the underlying models.
As adoption expands, a distinct services ecosystem is forming around OpenAI. Consulting and implementation providers are developing specialized expertise across the OpenAI API platform, ChatGPT Enterprise, Codex, agentic workflows, context engineering, and the integration of OpenAI capabilities into existing enterprise systems.
G2 created this category because general AI consulting services do not adequately distinguish between firms providing broad, model-agnostic AI guidance and those with demonstrated expertise deploying OpenAI technologies in production. Buyers pursuing an OpenAI initiative need a clearer way to identify providers that understand the platform’s architecture, capabilities, enterprise controls, development patterns, and rapidly evolving product ecosystem.
The need for specialized support becomes especially important as organizations move from pilots to scaled deployments. Providers may help enterprises identify and prioritize use cases, design AI-native workflows, build applications and agents, integrate proprietary data and systems, establish governance and evaluation practices, enable users, and translate technical deployments into measurable business outcomes.
This category also reflects the broader maturation of OpenAI’s enterprise ecosystem. As OpenAI becomes embedded within more business applications and workflows, implementation partners are playing a larger role in helping organizations bridge the gap between access to advanced models and the operational work required to deploy them successfully.
OpenAI Consulting Services gives buyers a dedicated place to find partners with validated OpenAI expertise and distinguishes an emerging provider ecosystem that would otherwise remain grouped within broader AI consulting categories.
- Bijou Barry, AI Principal Analyst
The top three providers in this category, ranked by popularity at the time of launch, are:
About this category: Cyber Risk Quantification is a new and rapidly growing category of software that allows IT and InfoSec professionals to communicate the necessities of ensuring their organizations create and maintain a robust cybersecurity posture to organizational stakeholders in a language everyone will understand: financially.
While the dangers and risks of poor security postures are intimately understood to legal and security teams for compliance and digital safety reasons, communicating with other areas within an organization through financial language has proven to better prioritize the purchasing of new software, faster protocol update reviews, and strengthen interdepartmental understanding of security practices and their importance.
Why it was introduced: “G2’s new Cyber Risk Quantification category captures this emerging market of software’s initial products, so buyers can effectively work within their organizations to substantially reduce lag time between advocating for security event preparedness and actually being prepared for those security threats.
Different teams within businesses, nonprofits, governmental agencies, or other types of organizations all understand their responsibilities through different languages, but financial risk remains a universal way for all parties to concisely understand the urgency required for maintaining and updating cybersecurity protocols.”
- Brandon Summers-Miller, Research Principal, Cybersecurity & Data Privacy
The top three products in this category, ranked by popularity at the time of launch, are:
On average, G2’s Market Research team adds 5–10 new categories per month, so we encourage you to review our research agenda and utilize the all-new G2.ai to leave a voice review on the software you recommend!
Come back next month to see additional new categories on G2! And, check out the new categories that went live last month.