AI agent disclosures in SEC filings, deployment confirmed
As of 2026-02-25, MERCADOLIBRE INC (MELI) disclosed AI-agent language in 1 SEC filing since 2026-02-25, including 2 excerpts describing its own deployment of AI agents (source: Agentic 10-K, from SEC EDGAR).
IT & security ops, Legal & compliance
None named
1 filing since 2025-01-01. All EDGAR filings (opens in a new tab)
Stage and function are keyword-rule labels, not the company’s words.
Sentences copied from the filings, whitespace normalised.
“We also deploy cross-ecosystem protections—including prompt-injection defenses, confidentiality controls for conversations and sensitive data, a strong focus in agentic identity management to ensure secure authentication and authorization across AI-driven interactions, and dedicated initiatives to identify and mitigate shadow AI usage across the organization.”
“We also use autonomous AI agents to support portions of our engineering work.”
“Additionally, the integration of autonomous AI agents into our workflows introduces heightened risks related to unpredictable system behavior and reduced human oversight, which could lead to significant operational errors, cybersecurity vulnerabilities, or inaccuracies.”
MercadoLibre uses autonomous AI agents for code reviews, documentation, incident review and migrations.
“We also use autonomous AI agents to support portions of our engineering work. These agents perform defined tasks such as contributing to code reviews, assisting with documentation, reviewing incidents, and handling certain migration and update activities across parts of our codebase. They operate as part of our broader development workflow and complement the work of our engineering teams.”
MercadoLibre notes risks from integrating autonomous AI agents including unpredictable behavior and reduced oversight.
“We are expanding our investment in AI/ML across the entire Company. This includes using generative AI and continuing to integrate AI capabilities into our products and services. While AI/ML can present significant benefits, it can also present risks and challenges to our business. Because AI/ML is a developing technology, legal frameworks for AI/ML governance are unsettled, quickly developing, and unpredictable. The use of AI/ML could lead to legal and regulatory investigations and enforcement actions, or may give rise to specific obligations, including required notices, consents and opt-outs, under various data privacy, protection and cybersecurity laws and regulations in a number of jurisdictions. Additionally, the integration of autonomous AI agents into our workflows introduces heightened risks related to unpredictable system behavior and reduced human oversight, which could lead to significant operational errors, cybersecurity vulnerabilities, or inaccuracies. Any failure to properly govern or control these autonomous agents could lead to regulatory non-compliance, potential legal liabilities and reputational harm arising from unauthorized or unintended agent actions. Some uses of AI/ML pose emerging ethical issues and present a number of risks that cannot be fully mitigated, including data sourcing, technology integration, bias in decision-making algorithms, discrimination or AI hallucinations, open source software issues, security challenges and the protection of perso”
MercadoLibre deploys AI-related cybersecurity protections including agentic identity management.
“Our risk management framework includes processes to manage cybersecurity risks associated with AI, such as enhanced governance of AI-related traffic, expanded automated scanning and blocking capabilities, and the building of comprehensive inventories that support continuous monitoring. We also deploy cross-ecosystem protections—including prompt-injection defenses, confidentiality controls for conversations and sensitive data, a strong focus in agentic identity management to ensure secure authentication and authorization across AI-driven interactions, and dedicated initiatives to identify and mitigate shadow AI usage across the organization.”