A Brazilian court’s order requiring Meta to demonstrate equal treatment of Lula and Bolsonaro raises a question that extends beyond the treatment of either political figure: what evidence should establish that a platform is acting impartially? The demand makes equal treatment something Meta must substantiate, rather than simply assert. Yet a meaningful assessment requires a distinction between applying consistent rules and producing matching results. Without that distinction, an accountability exercise could become a comparison of outcomes that says little about whether the underlying decisions were justified.

The order increases the penalty for noncompliance from 500,000 Brazilian reals ($100,000) to one million reals ($200,000). Those figures establish the financial stakes of failing to comply; they do not establish that unequal treatment has occurred. The central issue is the court’s demand for proof concerning Lula and Bolsonaro. The distinction matters: an order to demonstrate equal treatment is not, by itself, evidence resolving the question under examination. Nor does the larger penalty reveal what evidence would satisfy the court.

A demand for transparency also comes from CTRL+Z. CTRL+Z argues that Brazilian citizens are entitled to explanations when accounts are banned, posts disappear, reach diminishes or restrictions prevent commenting, reposting and saving material. Its position brings the experience of individual users into the accountability debate. Whereas the court’s order concerns equal treatment of Lula and Bolsonaro, CTRL+Z’s stated concerns encompass the reasons behind several kinds of platform restriction. The demands concern proof and explanations respectively, but the available material does not establish that they seek the same remedies.

Read more: Spain’s pension debate turns on the cost of retirement security

One way to read the court’s demand is as an assertion that impartiality should be capable of examination. If Meta is required to demonstrate equal treatment, a bare assurance would not answer the underlying evidentiary question. A useful assessment would ask whether comparable circumstances receive comparable treatment and whether differences have an intelligible basis. That is an analytical standard, not a description of requirements specified in the available account of the order. It highlights why the quality of the explanation matters as much as the existence of a response.

CTRL+Z’s argument suggests a complementary test: can affected users understand what happened to them? Its list of concerns distinguishes account access, the continued availability of posts, their reach and the ability to interact with content. These should not automatically be treated as interchangeable measures. An explanation for an account ban would not necessarily explain diminished reach, just as an explanation for a missing post would not necessarily account for an inability to save material. On that reading, transparency would need to identify the particular action being questioned.

An important analytical caution is that equal treatment cannot safely be inferred from equal outcomes alone. If the circumstances surrounding two decisions differ, different results could be consistent with impartial rules. Conversely, matching results would not necessarily establish that the decision-making process was fair. This is a counterargument about how to assess the order, not a position that can be attributed to Meta on the supplied evidence. No response from the company, Lula or Bolsonaro is included in the available material, so their explanations or objections cannot responsibly be supplied on their behalf.

Read more: Trump plans NIH women’s health research division after Melania Trump’s push

The financial figures also require restraint. The move from 500,000 Brazilian reals to one million reals makes noncompliance more costly under the order. It does not disclose the cost to Meta of assembling evidence, the adequacy of its records or the practical effect of the sanction. The dollar equivalents, $100,000 and $200,000, describe the same penalty amounts rather than separate enforcement measures. Reading the figures as proof of either effective accountability or excessive intervention would go beyond what they establish.

The competing considerations are therefore evidentiary as well as institutional. The court is demanding a demonstration of equal treatment; CTRL+Z is pressing for explanations of restrictions affecting citizens. In assessing those demands, a methodological concern arises: an assessment focused on superficial symmetry could misidentify both fair and unfair decisions. The available material supports scrutiny of how Meta treats the two figures, but it does not support a finding that either has been favoured. Maintaining that boundary is essential to a balanced reading of the dispute.

What comes next should be judged by the substance of any demonstration Meta provides, rather than by the penalty alone. Questions worth asking include whether its explanation distinguishes different restrictions, addresses comparable circumstances and gives reasons for divergent decisions. These are criteria for evaluating a response, not claims about evidence already submitted. The supplied account does not establish what Meta has produced or how the court would assess it. On the available evidence, the order represents a demand for accountability, not a completed assessment of platform impartiality.