A year after popularly elected judges took office, the experiment has neither collapsed nor delivered on its promises. What comes next will matter to every business with exposure to Mexico. By Juan Carlos Machorro, Santamarina y Steta | Mexico City | October 2026

On September 1, 2025, Mexico became the only country in the world to entrust the selection of its entire top judiciary, including all nine Supreme Court justices, to the ballot box. Twelve months on, the debate has coalesced around two clearly opposed narratives. The Court’s leadership and the governing coalition describe a judiciary that is functioning, autonomous and austere. Think tanks, judges’ associations, bar associations and most legal specialists see something else: lower productivity at the top, a predominance of procedural rulings, the loss of a professional career judiciary, lingering doubts about independence and an electoral model that the government itself has already had to amend.

The truth, as is often the case, lies in the details. And the most telling detail of the year is that, before its second phase could begin, the 2024 design was corrected. A constitutional amendment approved by the Senate on May 29, 2026 and published on June 2, 2026 postponed the next judicial election from 2027 to June 2028, reduced the number of candidacies, simplified ballots and introduced stricter screening. Critics read this as a tacit admission that the original design did not work as intended; supporters describe it as a technical adjustment. Either way, it frames the questions that lawyers and investors should now be asking.

The case for reform

Any fair assessment must begin with the reasons the reform was adopted. Its proponents argue that it was necessary to broaden access to justice, to end the excesses and privileges of the past and, above all, to root out nepotism in the federal judiciary. President Claudia Sheinbaum has presented the overhaul as part of the people’s fight against corruption, insisting that “nepotism is corruption in the Judiciary” and that, under the new model, any qualified law graduate may compete for a judgeship without depending on family connections. In his first annual report, Chief Justice Hugo Aguilar echoed the message: “privileges are over.”

The nepotism argument carries weight. The official narrative holds that of roughly 55,000 people working in the federal judiciary, about 48% were related to one another. The studies most often cited are narrower but still striking: a 2016–2017 internal review by the former Federal Judicial Council found that 501 of 1,031 judge and magistrate posts examined in 31 circuits (48.6%) had relatives working in the judiciary, with family networks reaching more than 7,000 public servants, and a later analysis put the share of employees with at least one relative on the payroll at about 49%. Critics reply that the problem has not disappeared; in August 2026, the federal judges’ association JUFED pointed to family ties within the new structures themselves.

What the first year has delivered

Measured by output, the picture depends on who is counting. Chief Justice Aguilar reports that the Court received 15,051 cases and resolved 13,231, or 88%, and that only 44 of the 1,479 matters inherited from its predecessor remain pending. “There is no backlog, there is no inefficiency,” he declared, while acknowledging that publication of statistics had been suspended “until the system improves.” Independent analyses are less sanguine. Integralia found that the new Court resolved 1,218 cases between September 2025 and February 2026, 37% fewer than its predecessor in the same period, and that seven in ten concluded on procedural grounds without any review of the merits. Expansión calculated an 18% decline over the full year, and litigators note that only 2,640 of the cases reported as resolved actually reached the Plenary, where substance is decided.

Equally significant is the Court’s relationship with the Executive. Integralia documented 196 matters in which the Federal Executive was a party; of the 64 decided on the merits, the Court sided with the government in 53. More than half of the 1,118 votes analyzed were unanimous. Supporters see a cohesive and balanced bench; critics see a weakened counterweight to political power.

An election that raised more questions than it answered

Much of the skepticism traces back to the June 1, 2025 election itself. Turnout stood at just 12.9%. Voters faced 881 posts across multiple ballots listing thousands of largely unknown candidates, a process the OAS Electoral Observation Mission described as “extremely complex” and “polarizing,” marked by a “great lack of knowledge” among the electorate, and one it recommended other countries in the region not replicate. The final count included 54.2 million null votes, and media reports of INE data placed abstention at about 87%.

The integrity of the vote was also called into question. Voting guides, known as acordeones, circulated widely despite precautionary measures that the electoral authority issued only two days before election day. The OAS observed that most of the top vote-getters had been promoted through them, and INE later found that winning candidates appeared on an average of 80% of documented guides. The electoral authority itself acknowledged polling stations where turnout reached or exceeded 100% of registered voters, as well as ballots without fold marks, and the election was held without a preliminary results program. The Konrad Adenauer Foundation summed up the exercise as hurried, flawed in design and execution, and low in legitimacy.

Persistent concerns and partial relief

For the legal community, the consequences extend well beyond the ballot. JUFED speaks of “the dismantling of the professional justice system,” as experienced judges were replaced by candidates with little or no judicial experience. The Barra Mexicana, Colegio de Abogados has warned that attempts to revisit final judgments issued by the former Supreme Court chambers would threaten res judicata, due process and the separation of powers. International observers, including the IBA, have flagged risks to amparo protection, to judicial review of tax and regulatory decisions and to the enforcement of arbitral awards, concerns that have become intertwined with the 2026 USMCA review. The new Judicial Discipline Tribunal, meanwhile, is widely seen as the institution whose conduct will reveal whether the system enforces accountability or pressures independent judges.

Some fears, however, have eased. The catastrophic scenarios forecast by certain critics have not materialized: courts operate, and digital files, random case assignment and online tracking have been introduced. Separating the next judicial vote from the 2027 midterms avoids overloading the electoral authority, and the 2026 amendment adds knowledge exams, background assessments, permanent training, first-year performance reviews and the possibility for the Supreme Court to sit in two sections.

The road to 2028

The debate has thus shifted from whether the model works to how it should be corrected before Mexicans return to the polls in June 2028, when the remaining federal judges and magistrates and the judiciaries of seventeen states will be renewed. The pending reform to the Organic Law of the Judiciary now offers a narrow window to go further than the constitutional amendment did. In our view, three priorities stand out.

First, a simpler and cleaner election. The judicial vote should be held on its own, apart from any other ballot, including a possible presidential recall vote. Each post should feature only a short list of candidates, with renewals staggered over several cycles rather than concentrated in a single day. Voting guides should be expressly prohibited and effectively sanctioned, each judge should be elected by the whole territory within his or her jurisdiction, and a preliminary results program should be restored.

Second, a far more demanding filter for candidates. Evaluation committees should be genuinely independent of the political branches, with participation from academia, bar associations and specialized organizations, and subject to verifiable conflict-of-interest rules. Selection by lottery should give way to merit-based rankings built on public rubrics. Candidates should meet minimum experience thresholds appropriate to each level, pass rigorous and eliminatory knowledge and case-resolution examinations, and undergo thorough integrity vetting, including scrutiny of any links to political parties or organized crime. A cooling-off period for party leaders and senior political officials, public hearings and full publication of candidate files would complete the picture.

Third, certainty for alternative dispute resolution. Consistent with legislative proposals to create specialized courts, Mexico would benefit from federal courts or chambers with exclusive competence over the recognition, enforcement and annulment of arbitral awards. Staffed by judges with proven expertise in commercial and international arbitration and applying the Commercial Code and the New York Convention consistently, such courts would shorten enforcement proceedings and give investors confidence that the outcome of arbitration will be honored.

What this means for business

Until these changes are adopted and tested in practice, companies operating in Mexico should anticipate a period of greater uncertainty in judicial review and in the enforcement of contractual rights. We have therefore recommended that our corporate and M&A clients review their portfolio of contractual relationships to identify where arbitration clauses make sense, with carefully chosen seats and institutional rules. In certain transactions, we have also suggested seeking guarantees or collateral over assets located outside Mexico, so that rights can be enforced without having to rely on the Mexican courts. Longer timelines in amparo, tax and regulatory litigation should be built into deal planning, and the Organic Law reform, the 2028 electoral rules, the Discipline Tribunal and the USMCA review all merit close monitoring.

A balance still open

Mexico’s judicial reform was justified as a remedy for limited access to justice, entrenched privilege and nepotism. Twelve months later, the evidence confirms neither the collapse predicted by its critics nor the transformation promised by its champions. The judiciary works, but with a thinner and more procedural output at the top, close alignment with the Executive and unresolved questions of independence. The 2028 election offers a second chance to get the design right. Whether Mexico seizes it will shape not only the credibility of its courts, but also the confidence of those who invest, contract and litigate there.

______________________________________________________________________________________________- This article is based on public sources, official communications and specialist reports available as of October 2026, including the Mexican Senate, the Official Gazette, the SCJN, INE, the OAS Electoral Observation Mission, Integralia, the Konrad Adenauer Foundation, JUFED, the Barra Mexicana, Colegio de Abogados, the IBA and Mexicanos Contra la Corrupción y la Impunidad. It is provided for general information and does not constitute legal advice.

Dow Jones Risk Journal

Publications

Parisa Salehi wrote an article in Dow Jones Risk Journal about what foreign investment in U.S. agricultural land means from an infrastructure and data control viewpoint.

"The security concern is real — but it is not adequately described by pointing to who owns the land," Parisa wrote. "It is better described by asking what any actor, of any origin, is doing with the agricultural infrastructure and data they control, and whether the U.S. has any capability to find out."

"State intelligence services have used commercial and agricultural cover to position collection assets near military installations and critical infrastructure," Parisa wrote. "This is an established tradecraft. Intelligence assessments across multiple agencies document how state actors routinely use commercially legitimate operations to establish proximity, access, and collection capability."

Click on the following link to read the full article: Dow Jones Risk Journal

Dow Jones Risk Journal reaches a global audience of risk and compliance professionals, as well as senior leaders who are responsible for assessing and mitigating risks.

In Italy, the use of AI in advertising already sits within a well-established framework requiring commercial communications to be transparent, truthful and non-misleading. These principles are enforced both under Italian consumer protection rules, reflecting EU law, and through the Italian advertising self-regulatory system, while the Italian Competition Authority (AGCM) may intervene against misleading commercial practices.

Against this background, the European Commission’s final Guidelines on Article 50 of the AI Act, adopted on 20 July 2026, provide important clarification on when the use of AI itself must be disclosed. The key point is that the mere use of artificial intelligence in the creation of advertising does not automatically trigger a disclosure obligation. The transparency obligations under Article 50 have applied since 2 August 2026 and, for deployers, are particularly relevant in the cases covered by paragraph 4, including deep fakes.

This does not, however, remove the need for a case-by-case assessment. Whether disclosure is required depends on the content actually disseminated, the way AI has been used, the degree of realism of the resulting content and, crucially, the perception it is likely to create for consumers in the relevant context.

In practice, disclosure will not always be required. AI may, for example, be used for minor technical adjustments, clearly fantastical or stylised settings, or to create an AI-generated background in which a real product is placed, provided that this does not alter consumers’ perception of the product, its characteristics, performance or use.

This approach has already been reflected in France by the ARPP – Autorité de Régulation Professionnelle de la Publicité, which published specific practical guidance on 3 August 2026. A similar distinction can be found in the recent WFA and ICAS guidance, which differentiates between uses of AI that affect consumers’ perception of the advertising message and uses that are merely technical, ancillary or decorative.

Where disclosure is required, it must be clear, distinguishable and understandable. Neither the AI Act nor the Commission Guidelines currently prescribe standard wording or a mandatory pictogram. Labels such as “AI-generated image”, “AI-generated video” or “AI-generated voice” are therefore examples rather than compulsory formulas.

The Italian advertising self-regulatory system is also considering the adoption of specific guidelines on the use of AI in commercial communications, including on appropriate labelling methods. We will keep you updated on the next developments in Italy.

For further information contact: Margherita Barié Head of Carnelutti IP Team

Read the article here Link