Some go-to-market (GTM) teams are deploying AI agents without being able to count which ones touch customer records or audit every action. In a 2026 survey of 157 B2B revenue, marketing and sales operations leaders, LeanData reported that 93% had at least one agent in production, while about one-third could not say how many agents touched their records and 30% found actions without an audit trail. Those findings point to an operating-control gap in this surveyed group—not proof that every GTM organization has uncontrolled agent sprawl.
What the survey says about agent visibility
LeanData’s 2026 report describes agent adoption alongside limited visibility among the 157 B2B revenue, marketing and sales operations leaders it surveyed. The report page says 93% had at least one agent in production and only 31% believed their infrastructure was ready. A separate article on the survey’s findings identifies May 2026 fieldwork and reports that about one-third of respondents could not say how many agents touched their records; 30% found actions taken without an audit trail. These are survey results published by LeanData, a GTM technology provider, rather than an independent census of all GTM teams. LeanData’s report and the survey coverage provide the respective details.
As an Amazon Associate I earn from qualifying purchases.
There is also a wider enterprise warning, though it should not be mistaken for a GTM-specific count. Gartner forecast in 2026 that the average global Fortune 500 enterprise would have more than 150,000 agents in use by 2028, up from fewer than 15 in 2025. Gartner also reported that only 13% of organizations thought they had the right AI-agent governance in place. The first figure is a forecast, and the second reflects organizations’ assessment of their own governance—not a measured count of agents in GTM teams. Gartner’s April 2026 release explains both figures.
Why one inventory is harder than it sounds
Agents can enter a revenue stack through several routes, not just a centrally approved agent platform. In the survey coverage, respondents reported AI features built into GTM tools (69%), custom applications using LLM APIs (62%) and agent platforms (46%). These categories can overlap: a team may use all three. A count limited to agents registered in a dedicated platform can therefore miss automation embedded in existing tools or built directly against an API. The survey article reports the routes and percentages.
#1 Best Overall
The practical question is not only “How many agents have we bought?” but “What software can act on a lead, contact, account, opportunity or campaign, under whose authority, and through which connection?” A useful inventory has to include sanctioned and less visible agents and distinguish the agent from the human or team accountable for it.
How agents amplify familiar GTM problems
Agents inherit the quality of the data, integrations and procedures they are given. LeanData’s 2026 survey found that 55% of respondents named data quality as a top AI challenge; 70% said data hygiene degraded execution. In the same survey, 45% cited bad data as a reason AI initiatives stall, 37% cited undocumented processes and 32% cited siloed teams. These are respondent-reported challenges, not independently measured rates across the industry. LeanData’s report describes the findings.
In a GTM workflow, a stale record or unclear handoff can become a coordination problem when several automated systems act on it. Survey examples include multiple tools or agents contacting the same prospect and a marketing sequence continuing while a sales representative is closing a deal. The remedy is not simply to add another agent: teams need clear customer-data definitions, documented routing and campaign rules, and agreement about which system or person owns each handoff.
The scale of adoption in marketing technology offers related context, but it is a different sample and should not be combined with the LeanData figures. Gartner’s 2025 survey of 413 marketing technology leaders, fielded June through August, found 81% were piloting or had implemented agent initiatives. Among martech leaders with agents in pilots or production, 45% said vendor-offered agent capabilities did not meet performance expectations. The results indicate that deployment and performance concerns can coexist; they do not establish that the same rates apply to revenue operations or to all GTM teams. Gartner’s October 2025 release gives the survey scope.
Build a workable inventory and control process
The following sequence synthesizes Gartner’s governance recommendations with the GTM data and workflow issues highlighted in LeanData’s survey. It is a practical framework, not a tested implementation recipe.
- Set creation and action rules. Define who may create or share agents, which connectors are permitted, and what actions an agent may take without human approval. Gartner recommends policies for agents and their connectors.
- Discover agents centrally. Record sanctioned agents and look for agents embedded in GTM products, custom API applications and other platforms. For each entry, capture its purpose, accountable owner, business workflow, connected systems and operating status. Gartner recommends a centralized inventory and categorization; ownership and purpose make that inventory useful to a GTM team.
- Give agents identifiable, limited access. Assign agent identities where possible, link each to an accountable operator and grant only the data and actions required for its role. Review access as workflows change, and retire redundant or obsolete agents. Gartner recommends identity and access controls, review and retirement.
- Govern the information agents use. Check data access, definitions and freshness, including whether an agent can act on outdated customer information or expose more than its task requires. Align the underlying sources of truth and document routing, campaign rules and handoffs. LeanData’s survey identifies poor data quality and undocumented processes as recurring obstacles.
- Monitor behavior and respond to exceptions. Watch for actions outside an agent’s intended scope, such as unexpected record changes or conflicting outreach, and define who can pause, correct or disable the agent. Gartner recommends behavior monitoring and remediation, alongside employee training.
- Keep useful activity records with privacy limits. Retain logs that can help attribute actions and investigate incidents. Decide who can inspect them, what activity is collected and how long it is retained; visibility can improve accountability while also concentrating power over employee and customer activity.
- Train users and maintain the process. Explain permitted uses, escalation routes and how employees can report an unlisted agent or unsafe behavior. Gartner recommends training and sharing responsible-use practices; assign someone to keep the inventory and policies current.
Choose visibility mechanisms for the job
A 2024 ACM FAccT research framework groups agent visibility into identifiers, real-time monitoring and activity logs. It is a conceptual framework, not a product comparison or proof that any single mechanism is sufficient. The paper also highlights privacy and the risks of concentrating oversight.
| Mechanism | What it helps answer | What it does not establish by itself |
|---|---|---|
| Agent identifiers | Which agent or identity is associated with an action. | Whether the action was appropriate, or whether every agent has been discovered. |
| Real-time monitoring | What an agent is doing while it operates, allowing teams to spot behavior that needs attention. | A complete history suitable for later attribution or investigation. |
| Activity logs | What actions were recorded for later review and accountability. | That unlogged actions did not occur, or that logs should be broadly accessible without privacy controls. |
These mechanisms serve different purposes: identifiers support attribution, monitoring helps teams react during activity, and logs support later review. A governance or monitoring product may help with discovery and oversight, but software alone cannot define sound permissions, repair source data, document handoffs or decide who is accountable. Gartner identifies AI TRiSM tools as one possible aid to discovering and categorizing agents, without endorsing a particular vendor in its release.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
What the evidence does—and does not—show
The clearest evidence for the title’s concern comes from LeanData’s survey: its respondents reported production use alongside difficulty counting agents that touched records and gaps in audit trails. Gartner’s forecast underscores the potential scale of enterprise agent use, while its marketing technology survey shows adoption and unmet expectations in a distinct respondent group. Together these sources make a case for treating inventory, permissions, process quality and monitoring as operating responsibilities—not for assuming every GTM team has lost control or that one product can restore it.
Quick Recap
Best Value
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
