Resolution 2/2022 covers specified micro and small enterprises, startups, private non-profits, natural persons, and unincorporated private entities acting as controllers or operators.
Qualification alone is not enough: revenue, economic-group, and high-risk exclusions can remove the differentiated treatment.
Use this guide to test whether a controller or operator is a under Resolution 2/2022, identify the available flexibilities, and document the exclusions and continuing LGPD duties.
1
Section 1
What should teams decide about Small Processing Agents under the Brazil LGPD?
Apply three checks before relying on Resolution 2/2022. First, confirm that the organization fits an eligible category in Article 2. Second, confirm that its gross revenue and, where relevant, the global revenue of its economic group do not exceed the limits incorporated by Article 3. Third, test whether its activity is . An agent that fails any applicable check cannot use the differentiated treatment for small processing agents.
For the incorporated revenue tests, a microenterprise has annual gross revenue up to R$360,000 and a small enterprise has revenue above R$360,000 and up to R$4.8 million under Complementary Law 123/2006. A startup must remain within the Complementary Law 182/2021 ceiling of R$16 million in the previous calendar year, or R$1,333,334 multiplied by the number of months of activity when it operated for less than 12 months. Use the applicable statutory period and include the global revenue of a de facto or de jure economic group.
High risk under Article 4 requires at least one general criterion together with at least one specific criterion. The general criteria are large-scale processing or processing that may significantly affect interests and fundamental rights. The specific criteria are emerging or innovative technology, monitoring publicly accessible areas, decisions based solely on automated processing, or use of sensitive data or data about children, adolescents, or older people.
Keep registration and corporate-form evidence for the claimed eligible category.
Record the revenue period, applicable statutory limit, group entities, group revenue, and calculation owner.
Assess scale, duration, frequency, geographic reach, possible denial of rights or services, and material, moral, discrimination, fraud, identity, image, and reputation harms.
If the ANPD requests proof of eligibility, Article 5 gives the agent 15 days to demonstrate that it meets Articles 2 and 3.
Which obligations are simplified, and which still apply?
The regime permits a simplified processing record, a simplified security policy, and specified differentiated periods. Qualifying agents are not required to appoint an encarregado, but an agent without one must maintain a communication channel for data subjects. Appointing an encarregado is treated as a governance practice, not as a condition of qualification.
The flexibilities do not waive the LGPD's principles, legal bases, transparency, data-subject rights, security, incident handling, other laws, contracts, or accountability. Security measures must still be essential and necessary for the actual privacy risk and the agent's circumstances.
Rights: provide accessible treatment information and a usable request channel; the regulation does not remove Article 18 rights.
Records: use the ANPD simplified processing-record model if suitable, but keep enough detail to show purposes, data, roles, sharing, retention, security, and rights handling.
Security: adopt necessary administrative and technical measures and keep a policy proportionate to structure, scale, volume, cost, and risk.
Timing: verify each differentiated period. The regulation does not create a blanket extension for every LGPD duty.
Which timing rules and exceptions need special care?
Article 14 grants double time for specified data-subject requests, the Article 19 complete declaration, incident communication under Resolution 15/2024, and information or records requested by the ANPD. Article 15 allows the simplified Article 19 confirmation or access declaration within up to 15 days rather than immediately. Read the exact provision before calculating a due date.
The ANPD may require a small agent to comply with an otherwise waived or flexible obligation after considering the nature or volume of processing and risks to data subjects. A prior eligibility decision should therefore be reopened when revenue, group structure, scale, technology, automation, monitored spaces, or affected populations change.
Do not use the small-agent regime for exclusively personal, non-economic household processing; that activity is outside the regulation and is addressed by the LGPD's scope exclusions.
Do not assume that non-profit status, a small headcount, or a supplier contract proves eligibility.
Example of high risk: large-scale use of sensitive health data meets a general criterion and a specific criterion. A small trial of innovative technology meets only a specific criterion unless it is also large scale or may significantly affect rights; document both sides instead of treating any single criterion as decisive.
High-risk agents may use the collective complaint negotiation option in Article 8, but that exception does not restore the other differentiated treatment.
Keep a dated eligibility memo and reassessment trigger beside every flexibility the organization uses.
What evidence should support the eligibility decision?
Maintain one controlled eligibility record that identifies the legal category, revenue and group tests, high-risk analysis, each flexibility used, the supporting evidence, owner, approval date, and reassessment conditions. Link it to the processing inventory so reviewers can see whether later processing changed the conclusion.
A controller and an operator may each need their own assessment. Role labels do not decide eligibility, and qualifying as a does not change whether an entity is a controller or operator for a particular activity.
Corporate evidence: registration, legal form, startup qualification where used, revenue records, and economic-group analysis.
Processing evidence: inventory, scale measures, people and data categories, automation, technologies, locations monitored, possible harms, and controls.