Can Audio Deepfakes Lead to Cyber Libel and Fraud Charges?
Introduction
Artificial voice clones can make a recording appear to come from a corporate executive, director, auditor, or other trusted person. When manipulated audio is shared online to damage reputation, obtain money, or induce employees or investors to act, the conduct may give rise to criminal liability under Philippine law.
The applicable charge depends on the content of the recording, the manner of distribution, the offender’s intent, the resulting harm, and whether the offender personally created or originally posted the material. A mere recipient who reacts to or forwards an online post is not treated in the same manner as its original author for purposes of cyber libel.
When Manipulated Audio May Constitute Cyber Libel
Libel consists of a public and malicious imputation of a crime, vice, defect, act, omission, condition, status, or circumstance tending to cause the dishonor, discredit, or contempt of a natural or juridical person. The imputation must generally be defamatory, published, identifiable as referring to the offended party, and attended by malice, subject to recognized legal exceptions.
When the defamatory material is communicated through a computer system or online platform, it may constitute cyber libel under Section 4(c)(4) of R.A. No. 10175, in relation to Articles 353 and 355 of the Revised Penal Code. The Supreme Court has recognized that cyber libel is not an entirely new felony; it is libel committed through a computer system, with the use of information and communications technology increasing the penalty by one degree under Section 6 of R.A. No. 10175 (“Causing v. People of the Philippines, et al.,” G.R. No. 258524, 2023).
Accordingly, a manipulated recording may support a cyber-libel complaint if it falsely makes it appear that a corporate executive committed fraud, accepted a bribe, misused corporate funds, deceived investors, or engaged in another act that tends to damage the executive’s reputation. A corporation may likewise be the offended juridical person where the recording imputes misconduct to the company or tends to discredit its business standing.
Elements That Must Be Established
A complainant should be prepared to establish the following circumstances:
- Defamatory imputation. The audio, accompanying caption, or surrounding message must impute a crime, vice, defect, or other discreditable circumstance.
- Identifiable offended party. The executive or corporation must be identifiable either expressly or through circumstances known to the recipients.
- Publication. The material must have been communicated to at least one person other than the offended party.
- Malice. Malice may be presumed in a defamatory imputation that is not privileged, although the accused may invoke applicable defenses recognized by law.
- Use of a computer system. The recording or accompanying communication must have been posted, transmitted, or otherwise disseminated through an online or computer-based system.
The fact that an audio recording is artificial or manipulated does not, by itself, establish cyber libel. The prosecution must still prove the defamatory character of the communication, the identity of the person or entity targeted, publication, and the required criminal intent or malice.
Original Author Versus Persons Who Forward the Recording
Section 4(c)(4) of R.A. No. 10175 is valid and constitutional as applied to the original author of the online libel. The Supreme Court declared the provision unconstitutional as applied to persons who merely receive the post and react to it. The Court also invalidated the application of aiding-and-abetting and attempt provisions to online libel in the circumstances addressed in the decision (“Disini, Jr., et al. v. The Secretary of Justice, et al.,” G.R. No. 203335, 2014).
This distinction is important in investigating deepfake incidents. Evidence should identify who generated the audio, who first uploaded or transmitted it, who added the defamatory narration or caption, and whether later participants merely received or reacted to the post or independently republished it.
Prescriptive Period for Cyber Libel
Cyber libel prescribes in one year from discovery by the offended party, the authorities, or their agents, applying Articles 90 and 91 of the Revised Penal Code. The period does not automatically begin on the date the recording was first uploaded or published (“Causing v. People,” G.R. No. 258524, 2026).
Because discovery is determinative, complainants should document when and how the recording was discovered, who discovered it, the platform or account involved, and when the material was preserved. Delay may affect the timeliness of the complaint and the availability of digital evidence.
When the Conduct May Constitute Computer-Related Fraud
Manipulated audio may also support a computer-related fraud charge when it is used as part of a scheme to obtain money, property, access, or another benefit. Section 4(b)(2) of R.A. No. 10175 covers the unauthorized input, alteration, or deletion of computer data or a computer program, or interference with the functioning of a computer system, when the conduct causes damage and is attended by fraudulent intent.
The statutory definition requires more than the circulation of a false recording. The prosecution must generally connect the accused to an unauthorized computer-related act, fraudulent intent, and resulting damage. If no damage has yet been caused, the penalty imposable under the provision is one degree lower (Cybercrime Prevention Act of 2012).
For example, a deepfake voice message may be relevant to computer-related fraud if it falsely appears to direct an employee to transfer company funds, disclose banking credentials, change a supplier’s payment account, or approve a transaction. The audio is then evidence of the fraudulent scheme; the charge must still be supported by proof of the unauthorized computer-related act and actual damage or the absence of damage, as applicable.
Possible Fraud Scenarios Involving Executives
Common scenarios include the following:
- A fake voice message allegedly instructs an employee to transfer funds to a new bank account.
- A recording falsely portrays a chief executive as authorizing the release of confidential investor information.
- An artificial audio clip is circulated to investors to induce them to purchase shares, subscribe to a product, or release money.
- A fabricated conversation is sent to a finance officer to obtain passwords, one-time passwords, or access to a corporate system.
- A manipulated recording is posted publicly to depress confidence in a company and induce a sale or other financial transaction.
The legal characterization will depend on the evidence. A recording that merely insults an executive may raise a defamation issue. A recording that induces a transfer of money or causes unauthorized alteration of computer data may additionally support a fraud theory.
Cyber Libel and Fraud May Involve Different Injuries
Cyber libel primarily concerns reputational injury caused by defamatory publication. Computer-related fraud concerns unauthorized computer-related conduct carried out with fraudulent intent that causes damage. One incident may contain facts relevant to both offenses, but each offense requires its own statutory elements.
The prosecution should therefore avoid treating the existence of a deepfake as conclusive proof of every possible cybercrime. The complaint should separately identify the defamatory statements, the recipients, the publication method, the fraudulent representation, the computer-related act, the financial or system damage, and the evidence linking the accused to each act.
Evidence to Preserve
Deepfake cases are highly dependent on digital evidence. The offended corporation or executive should preserve the original file and not rely only on a screen recording or a forwarded copy.
- Original digital file. Preserve the audio in its original format, including metadata where available.
- Platform evidence. Record the account name, profile URL, post identifier, date, time, recipients, and method of transmission.
- Message history. Preserve e-mails, chats, captions, voice notes, and replies surrounding the audio.
- Financial records. Document attempted or completed transfers, altered payment instructions, unauthorized withdrawals, and related losses.
- Technical examination. Obtain a forensic assessment addressing manipulation, source files, editing artifacts, voice-cloning indicators, and the relationship between the file and the accused’s devices or accounts.
- Witness statements. Obtain statements from the person who received the message, the executive whose voice was imitated, and employees who acted or nearly acted on the instruction.
Evidence should be preserved in a manner that can establish authenticity, integrity, chain of custody, and the connection between the file, account, device, and accused person. A screenshot may help show publication, but it may not by itself establish authorship or prove that the recording was manipulated.
Relationship With Personal Data and Privacy Rules
A voice recording may involve personal information when it relates to an identifiable individual. Processing or disclosure issues may therefore arise under the Data Privacy Act of 2012, particularly where the recording is combined with identifying information or disseminated without lawful authority.
However, a privacy complaint does not automatically establish cyber libel or computer-related fraud. Under Section 25 of R.A. No. 10173, unauthorized processing requires proof that personal or sensitive personal information was processed without the data subject’s consent or without authority under the law. The National Privacy Commission has described these as cumulative requirements that must be established by substantial evidence (NPC 19-1429, 2024).
Where the recording contains false or unwarranted personal information and is disclosed with malice or bad faith by a personal information controller, processor, or covered personnel, Section 31 of R.A. No. 10173 may also become relevant. The required circumstances include the covered status of the perpetrator, disclosure of personal or sensitive personal information, malice or bad faith, and information that is unwarranted or false (NPC 19-1438, 2025).
Online Defamation Before the Cybercrime Prevention Act
The timing of the publication matters. The Supreme Court ruled that, before the enactment of R.A. No. 10175, online statements could not simply be treated as libel under the phrase “similar means” in Article 355 of the Revised Penal Code. The Court applied the principle of noscitur a sociis and held that the statutory list did not then include computer systems or the internet (“Peñalosa v. Ocampo, Jr.,” G.R. No. 230299, 2023).
For present conduct, however, online publication is expressly addressed by Section 4(c)(4) of R.A. No. 10175, subject to the constitutional limitations recognized in Disini.
Recommended Corporate Response
A company confronted with a manipulated executive recording should first prevent further financial or operational harm. It should temporarily verify payment instructions, freeze suspicious transactions where possible, notify banks and relevant service providers, and warn employees through an authenticated communication channel.
The company should then conduct a controlled evidence-preservation process. It should avoid repeatedly forwarding the recording, because unnecessary republication may increase reputational harm and complicate the assessment of who created and disseminated the material.
A complaint should identify the intended victim, the original or earliest known source, the precise defamatory or fraudulent representations, the recipients, the computer systems used, and the actual damage. It should also distinguish the acts of the original author from those of persons who merely received, reacted to, or independently republished the material.
Conclusion
Sharing a manipulated audio deepfake of a corporate executive may result in cyber libel liability when the material is defamatory, identifies the executive or corporation, is published online, and is attributable to the original author or responsible disseminator. It may also support computer-related fraud charges when the recording forms part of an unauthorized computer-based scheme intended to obtain money or another benefit and causes damage.
The decisive issues are not simply whether the voice was artificial or whether the material went viral. Investigators and counsel must establish authorship, publication, defamatory content, fraudulent intent, unauthorized computer-related conduct, actual damage, and the applicable prescriptive period. Prompt preservation of the original digital evidence and immediate verification of financial instructions are therefore essential.
About Nicolas and De Vega Law Offices
Nicolas and de Vega Law Offices is a full-service law firm in the Philippines. You may visit us at the 16th Flr., Suite 1607 AIC Burgundy Empire Tower, ADB Ave., Ortigas Center, 1605 Pasig City, Metro Manila, Philippines. You may also call us at +632 84706126, +632 84706130, +632 84016392 or e-mail us at [email protected]. Visit our website https://ndvlaw.com.

