Peru - Penal Code AI Modifications (10525/2024)

Overview

Project 10525/2024 was presented as a parliamentary initiative aimed at adapting Peru’s criminal law to the new threats posed by artificial intelligence (AI). It was registered in March 2025 and incorporated into a consolidated committee text (dictamen) that collects and harmonizes four distinct projects (6573/2023-CR, 7072/2023-CR, 8746/2024-CR and 10525/2024-CR). The approved committee text proposes explicit amendments to multiple Penal Code provisions (articles cited in the project title) and to the Law No. 30096 (Law on Computer Crimes), introducing the “use of AI” as an express aggravating circumstance that allows the judge to increase custodial penalties up to one third above the statutory maximum when the crime was committed through or materially facilitated by AI. The Plenary approved the committee dictamen in April 2025 and the text was subsequently delivered to the Executive; the Congress press release and the official legislative record document the committee vote and plenary outcome. For the official Congressional announcement, see Congress press release (16 Apr 2025).

Definitions

The committee text sets out or relies on working definitions intended for penal application: "artificial intelligence" (broadly, systems that perform tasks that would otherwise require human intelligence), "content multimedia generated by AI" and "deepfakes" (multimedia falsifications produced or assisted by algorithmic techniques). The draft clarifies that aggravation applies when an offense is committed by use of AI or when AI materially facilitates planning, execution, dissemination, or concealment of illegal acts — including automated generation of pornographic content, falsified audiovisual material used for reputational harm, or algorithmic manipulation to defraud victims. Although some sponsor texts included more technical definitions, the final committee text prioritizes functional definitions oriented to criminal legal effects (i.e., whether the technology increased the capacity to commit or conceal the crime or to multiply harm).

Governance and Institutional Framework

The legislative package remains primarily within the penal enforcement domain. It assigns operational responsibilities to judicial and prosecutorial authorities for application and interpretation of the aggravating factor. The dictamen explicitly instructs the Executive — namely the Presidency of the Council of Ministers (PCM) and the Ministry of Justice and Human Rights — to issue the additional implementing norms necessary to apply the amendments within 60 calendar days from entry into force. Operational enforcement will therefore require coordination across the Office of the Public Prosecutor (Ministerio Público), judicial tribunals (Judiciary), the National Police (PNP) and sectoral regulators where relevant (e.g., communications regulator). The committee and plenary debates recognized the need to strengthen investigative capacity in digital forensics, evidence preservation and cooperation with platforms and foreign authorities to trace AI-assisted wrongdoing; those needs were referenced in the committee instructions to the Executive for regulatory and capacity-building measures. See the dictamen and congressional diary for institutional references: Diario de Debates (14 Apr 2025) and the autógrafas registry: Autógrafas enviadas (Apr 2025).

Key Focus Areas

The principal focus areas of the draft legislation are (1) aggravation of criminal liability where AI is used to commit, enable or amplify offenses; (2) targeted protection of particularly vulnerable legal goods such as sexual freedom (including protections against AI-based sexual exploitation and child sexual exploitation), honor and reputation (defamation via deepfakes), and intellectual property rights (AI-generated works and fraudulent appropriation); and (3) procedural and implementation measures to ensure the law can be applied (e.g., executive regulation, evidence preservation, and criminal dosimetry adjustments). The legislative text is limited in scope compared with broad AI governance bills: it does not create an administrative pro‑algorithmic conformity regime or registration obligations for developers, nor does it impose explicit transparency or audit requirements on private-sector AI model providers in its penal text; rather, it strengthens criminal consequences where AI materially increases the gravity of crime. The committee and Plenary emphasized public policy rationales: closing gaps in the Penal Code, deterring misuse of AI for large-scale harm, and protecting rights (privacy, IP, sexual autonomy) that can be uniquely threatened by synthetic content. The committee also incorporated language to preserve judicial discretion so that aggravation attaches only where AI meaningfully contributed to the offense’s severity or reach.

Implementation Framework

Implementation depends on (a) executive-level secondary rules to operationalize definitions and evidentiary procedures (mandated within 60 days of entry into force), (b) capacity-building for prosecutors, forensic analysts and judges to assess AI involvement in criminal acts, and (c) procedural tools to preserve and obtain logs, datasets and platform records needed to prove AI use. The dictamen instructs the PCM and the Ministry of Justice to issue implementing norms; these norms are expected to address procedural cooperation with online platforms, chain-of-custody for AI-generated evidence, protocols for collection of metadata and hashing of multimedia, and inter-agency coordination. Because the primary instrument is penal law, the Executive’s implementing regulations are the channel for technical and operational detail (e.g., standards for forensic authentication of deepfakes), with the Public Ministry and Judiciary responsible for practical application.

Monitoring and Evaluation

The bill itself does not create a dedicated monitoring body or an explicit evaluation mechanism; monitoring is expected to occur through ordinary criminal statistics, prosecutorial tracking, and inter-institutional coordination mandated by the Executive regulations. The congressional dictamen expects relevant ministries to report on implementation needs and results; in practice, monitoring will depend on statistical capture of AI-linked prosecutions, creation of specialized prosecutorial units or trainings, and periodic review by judicial councils or the PCM. Civil society, victims’ organizations and academic groups were cited in debates as potential contributors to post-enactment evaluation and recommendations. The absence of an explicit administrative supervisory agency makes post-enactment evaluation reliant on existing institutions and ad-hoc coordination instruments.

Penalties, Liability, and Appeals

Under the approved committee text, the main change is an express aggravating circumstance: when a crime is committed by use of AI or a similar technology, the judge may elevate the imposed custodial sentence by up to one third above the statutory maximum for the base offense. This mechanism affects sentencing/dosimetry (penalty ranges) rather than creating new substantive criminal offenses; it is therefore a relative intensification of penalties in qualifying cases. The draft preserves ordinary criminal defenses and due process; accused persons retain the right to appeal sentencing enhancements. The dictamen states that aggravation should be applied consistently with proportionality and the principles of criminal law; the courts will be required to make specific factual findings on the role of AI that justify aggravation. The final text sent to the Executive (autógrafa) formalizes the decision of the Congress to transmit the approved law for the Executive’s sanction and publication.

Relationship to Other Instruments

The measure amends the Penal Code (Decreto Legislativo 635) and the Law 30096 (Law on Computer Crimes), and it is explicitly linked to intellectual property protections mentioned in sponsor expositions. It operates alongside other sectoral AI initiatives (education curriculum inclusion, administrative guidelines) that were under discussion in parallel parliamentary committees. Because the instrument is penal in nature, it complements — but does not replace — administrative or civil remedies (for example, copyright, data protection, consumer protection or platform governance laws). The dictamen was drafted as a harmonizing instrument that merges several parliamentary proposals into a single penal-oriented response to AI-related harms.

International Alignment

The committee record explicitly references comparative concerns and emerging international guidance on AI-related harms (for instance, deepfake regulation, child sexual exploitation frameworks and copyright guidance addressing AI-generated works). While the bill is a domestic criminal-law instrument, it is intended to align with international obligations to prevent exploitation, child sexual abuse, and cross-border cybercrime. The executive implementing regulations will likely define mechanisms for international cooperation and mutual legal assistance when AI-assisted offenses implicate foreign infrastructure or cross-border platforms. The legislative text’s focus on aggravation mirrors approaches in other jurisdictions that intensify penalties where technologies increase harm or make crimes especially intrusive.

Implementation Timeline

Sources and References