#FactCheck- AI-Generated Deepfake Falsely Shows Lt Gen Devendra Sharma Speaking About Missing Aviators in Operation Sindoor

Introduction

A Reuters investigation has uncovered an elephant in the room regarding Meta Platforms' internal measures to address online fraud and illicit advertising. The confidential documents that Reuters reviewed disclosed that Meta was planning to generate approximately 10% of its 2024 revenue, i.e., USD 16 billion, from ads related to scams and prohibited goods. The findings point out a disturbing paradox: on the one hand, Meta is a vocal advocate for digital safety and platform integrity, while on the other hand, the internal logs of the company indicate the existence of a very large area allowing the shunning of fraudulent advertisement activities that exploit users throughout the world. 

‍

The Scale of the Problem

Internal Meta projections show that its platforms, Facebook, Instagram, and WhatsApp, are displaying a staggering 15 billion scam ads per day combined. The advertisements include deceitful e-commerce promotions, fake investment schemes, counterfeit medical products, and unlicensed gambling platforms.

Meta has developed sophisticated detection tools, but even then, the system does not catch the advertisers until they are 95% certain to be fraudsters. By having at least that threshold for removing an ad, the company is unlikely to lose much money. As a result, instead of turning the fraud adjacent advertisers down, it charges them higher ad rates, which is the strategy they call “penalty bids” internally. 

‍

Internal Acknowledgements & Business Dependence

Internal documents that date between 2021 and 2025 reveal that the financial, safety, and lobbying divisions of Meta were cognizant of the enormity of revenues generated from scams. One of the 2025 strategic papers even describes this revenue source as "violating revenue," which implies that it includes ads that are against Meta's policies regarding scams, gambling, sexual services, and misleading healthcare products.

‍

The company's top executives consider the cost-benefit scenario of stricter enforcement. According to a 2024 internal projection, Meta's half-yearly earnings from high-risk scam ads were estimated at USD 3.5 billion, whereas regulatory fines for such violations would not exceed USD 1 billion, thus making it a tolerable trade-off from a commercial viewpoint. At the same time, the company intends to scale down scam ad revenue gradually, thus from 10.1% in 2024 to 7.3% by 2025, and 6% by 2026; however, the documents also reveal a planned slowdown in enforcement to avoid "abrupt reductions" that could affect business forecasts.

‍

Algorithmic Amplification of Scams

One of the most alarming situations is the fact that Meta's own advertising algorithms amplify scam content. It has been reported that users who click on fraudulent ads are more likely to see other similar ads, as the platform's personalisation engine assumes user "interest."

This scenario creates a self-reinforcing feedback loop where the user engagement with scam content dictates the amount of such content being displayed. Thus, a digital environment is created which encourages deceptive engagement and consequently, user trust is eroded and systemic risk is amplified.

An internal presentation in May 2025 was said to put a number on how deeply the platform's ad ecosystem was intertwined with the global fraud economy, estimating that one-third of the scams that succeeded in the U.S. were due to advertising on Meta's platforms.

‍

Regulatory & Legal Implications

The disclosures arrived at the same time as the US and UK governments started to closely check the company's activities more than ever before.

• The U.S. Securities and Exchange Commission (SEC) is said to be looking into whether Meta has had any part in the promotion of fraudulent financial ads.
• The UK’s Financial Conduct Authority (FCA) found that Meta’s platforms were the main sources of scams related to online payments and claimed that the amount of money lost was more than all the other social platforms combined in 2023.

Meta’s spokesperson, Andy Stone, at first denied the accusations, stating that the figures mentioned in the leak were “rough and overly-inclusive”; nevertheless, he conceded that the company’s consistent efforts toward enforcement had negatively impacted revenue and would continue to do so.

‍

Operational Challenges & Policy Gaps

The internal documents also reveal the weaknesses in Meta's day-to-day operations when it comes to the implementation of its own policies.

• Because of the large number of employees laid off in 2023, the whole department that dealt with advertiser-brand impersonation was said to have been dissolved.
• Scam ads were categorised as a "low severity" issue, which was more of a "bad user experience" than a critical security risk.
• At the end of 2023, users were submitting around 100,000 legitimate scam reports per week, of which Meta dismissed or rejected 96%.

‍

Human Impact: When Fraud Becomes Personal

The financial and ethical issues have tangible human consequences. The Reuters investigation documented multiple cases of individuals defrauded through hijacked Meta accounts.

One striking example involves a Canadian Air Force recruiter, whose hacked Facebook account was used to promote fake cryptocurrency schemes. Despite over a hundred user reports, Meta failed to act for weeks, during which several victims, including military colleagues, lost tens of thousands of dollars.

The case underscores not just platform negligence, but also the difficulty of law enforcement collaboration. Canadian authorities confirmed that funds traced to Nigerian accounts could not be recovered due to jurisdictional barriers, a recurring issue in transnational cyber fraud.

‍

Ethical and Cybersecurity Implications

The research has questioned extremely important things at least from the perspective of cyber policy:

1. Platform Accountability: Meta, by its practice, is giving more importance to the monetary aspect rather than the truth, and in this way, it is going against the principles of responsible digital governance.
2. Transparency in Ad Ecosystems: The lack of transparency in digital advertising systems makes it very easy for dishonest actors to use automated processes with very little supervision.
3. Algorithmic Responsibility: The use of algorithms that impact the visibility of misleading content and targeting can be considered the direct involvement of the algorithms in the fraud.
4. Regulatory Harmonisation: The presence of different and disconnected enforcement frameworks across jurisdictions is a drawback to the efforts in dealing with cross-border cybercrime.
5. Public Trust: Users’ trust in the digital world is mainly dependent on the safety level they see and the accountability of the companies.

‍

Conclusion

Meta’s records show a very unpleasant mix of profit, laxity, and failure in the policy area concerning scam-related ads. The platform’s readiness to accept and even profit from fraudulent players, though admitting the damage they cause, calls for an immediate global rethinking of advertising ethics, regulatory enforcement, and algorithmic transparency.

With the expansion of its AI-driven operations and advertising networks, protecting the users of Meta must evolve from being just a public relations goal to being a core business necessity, thus requiring verifiable accountability measures, independent audits, and regulatory oversight. It is an undeniable fact that there are billions of users who count on Meta’s platforms for their right to digital safety, which is why this right must be respected and enforced rather than becoming optional.

‍

References

• https://www.reuters.com/investigations/meta-is-earning-fortune-deluge-fraudulent-ads-documents-show-2025-11-06/?utm_source=chatgpt.com
• https://www.indiatoday.in/technology/news/story/leaked-docs-claim-meta-made-16-billion-from-scam-ads-even-after-deleting-134-million-of-them-2815183-2025-11-07

‍

Procedural History: 

‍

The case started with a 2011 Madras High Court ruling that included the appellant’s personal information. In the case discussed, the court decided in 2024, the appellant went to the Madurai Bench of the Madras High Court to request that his name and other identifying information from that previous ruling be redacted. He argued that his right to privacy under Article 21 of the Indian Constitution was violated by the ongoing release of such private information into the public arena. He claimed that the revelation had hurt him in real ways, such as having his application for an Australian visa denied. Therefore, without compromising the ideals of open justice, the current procedures aimed to have the court recognize a person’s “Right to be Forgotten” within a broader framework of privacy and data protection.

‍

Background and Factual Matrix

‍

The appellant was charged under Sections 417 and 376 of the IPC. The trial court convicted him in 201, but later, the High Court in 2014 fully, completely and unconditionally acquitted him, which was not based on the benefit of doubt. Following the acquittal, he remarried and has three children. The judgment of both the High Court and the Trial Court has personal and intimate details about him. Being available in the public domain has caused him significant repercussions, as he was denied a visa to travel to Australia by authorities, citing the criminal cases. The appellant has filed a plea seeking a mandamus directing the Registrar General, Additional Registrar General, and Registrar (IT-Statistics) as R1, R2, R3 to redact his name and other identities from the acquittal judgment. He has sought a direction from Ikanoon Software Development Private Limited (R4) to reflect the redaction in its publication. 

‍

Issue 

‍

1. Whether a writ of mandamus can lie against a High Court for redaction of personal details from its own judgment, or does such a prayer tantamount to a High Court issuing a writ against itself?
2. Whether the High Court, being a Court of Record under Article 215 of the Indian Constitution, is entitled to preserve its record for perpetuity in its original form without any modification or redaction?
3. Whether the ‘Right to be Forgotten' can be recognised and enforced in the absence of a specific statutory provision or Supreme Court direction, given that it constitutes an exception to the fundamental principle of open courts and open justice?

‍

Adjudication and Reasoning

The division bench has allowed the Writ appeal and granted the following relief:

1. R4 directed to take down the judgment in Crl.A. (MD) No.321 of 2011 dated 30.04.2014 forthwith.
2. R1 to R3 directed to redact the name and other details of the Writ Petitioner relating to his identity from the judgment dated 30.04.2014 in Crl.A.(MD) No. 321 of 2011 and ensure that only the redacted judgment is available for publication or for uploading.

‍

Rule 

‍

1. Courts have a wide discretion in deciding whether to allow redaction or not. Such discretion can either be granted at the request of the party seeking redaction or, in appropriate cases, even suo moto by the court.
2. The accused who have earned full, complete and unconditional acquittal without any benefit of doubt have a legitimate claim to move forward for redaction of personal information.
3. The open Court doesn’t require absolute disclosure of all personal information, and the courts, while deciding the concern of privacy and the right to ensure that in litigations to leave behind parts of their past which are no longer relevant, have to balance the concept of open Court on the one hand and privacy concerns of a citizen on the other.
4. As the High Court is the repository of a wide range of information and is entitled to preserve the original record in perpetuity. However, without diluting the sanctity of the original record, the public reflection of that record can be moderated to preserve the privacy of the person to whom that record pertains.

‍

Reasoning 

‍

1. Drawing on the judgment K.S. Puttaswamy v. Union of India, the court found Article 21 to protect not only informational privacy but also the "right to be forgotten," which gives individuals the right to request the deletion of any personal data when there is no longer any legitimate public interest in retaining such information. Such irreparable reputational damage is thus an infringement on constitutional privacy that demands judicial redaction. 
2. The court rejected the argument that a writ against its own order is impermissible, drawing a distinction between challenging the legal correctness of a judgment and seeking redaction of personal information. Allowing redaction will not question the validity of the judgment; rather, it will simply change its public appearance to ensure privacy.
3. Since a High Court is a Court of Record with an obligation to preserve its judgments in their unaltered form forever, the court held here that such internal maintenance of complete records was not incompatible with the issuance of a redacted public version. Institutional integrity is maintained when the original kept in the archives is supplemented with a public version that masks the privacy areas.
4. Open justice principles work to establish transparency, accountability, and public confidence, but these are not absolute. The court took a proportionality stance: personal identifiers, where they neither educate nor have precedential value and continue to inflict harm, may be expunged without affecting the established legal principles of judgment.
5. Although the DPDP Act exempts courts from several statutory obligations, the court held that it can, by virtue of its inherent discretion, protect personal data, and in so doing, exercise that power without the need for any legislative command. Traditionally the Madras High Court rules provide for the possibility of restriction of certified copies, thus establishing redaction as feasible both legally and administratively.

‍

Over the last decade, battlefields have percolated from mountains, deserts, jungles, seas, and the skies into the invisible networks of code and cables. Cyberwarfare is no longer a distant possibility but today’s reality. The cyberattacks of Estonia in 2007, the crippling of Iran’s nuclear program by the Stuxnet virus, the SolarWinds and Colonial Pipeline breaches in recent years have proved one thing: that nations can now paralyze economies and infrastructures without firing a bullet.  Cyber operations now fall beyond the traditional threshold of war, allowing aggressors to exploit the grey zone where full-scale retaliation may be unlikely. 

At the same time, this ambiguity has also given rise to the concept of cyber deterrence. It is a concept that has been borrowed from the nuclear strategies during the Cold War era and has been adapted to the digital age. At the core, cyber deterrence seeks to alter the adversary’s cost-benefit calculation that makes attacks either too costly or pointless to pursue. While power blocs like the US, Russia, and China continue to build up their cyber arsenals, smaller nations can hold unique advantages, most importantly in terms of their resilience, if not firepower.   

‍

Understanding the concept of Cyber Deterrence

‍

Deterrence, in its classic sense, is about preventing action through the fear of consequences. It usually manifests in four mechanisms as follows:

• Punishment by threatening to impose costs on attackers, whether by counter-attacks, economic sanctions, or even conventional forces. 
• Denial of attacks by making them futile through hardened defences, and ensuring the systems to resist, recover, and continue to function. 
• Entanglement by leveraging interdependence in trade, finance, and technology to make attacks costly for both attackers and defenders. 
• Norms can also help shape behaviour by stigmatizing reckless cyber actions by imposing reputational costs that can exceed any gains. 

However, great powers have always emphasized the importance of punishment as a tool to showcase their power by employing offensive cyber arsenals to instill psychological pressure on their rivals. Yet in cyberspace, punishment has inherent flaws. 

‍

The Advantage of Asymmetry 

‍

For small states,  smaller geographical size can be utilised as a benefit. Three advantages of this exist, such as:

• With fewer critical infrastructures to protect,  resources can be concentrated. For example, Denmark, with a modest population of $40 million cyber budget, is considered to be among the most cyber-secure nations, despite receiving billions of US spending.  
• Smaller bureaucracies enable faster response. The centralised cyber command of Singapore allows it to ensure a rapid coordination between the government and the private sector. 
• Smaller countries with lesser populations can foster a higher public awareness and participation in cyber hygiene by amplifying national resilience.  

In short, defending a small digital fortress can be easier than securing a sprawling empire of interconnected systems. 

‍

Lessons from Estonia and Singapore 

‍

The 2007 crisis of Estonia remains a case study of cyber resilience. Although its government, bank, and media were targeted in offline mode, Estonia emerged stronger by investing heavily in cyber defense mechanisms. Another effort in this case stood was with the hosting of NATO’s Cooperative Cyber Defence Centre of Excellence to build one of the world’s most resilient e-governance models. 

Singapore is another case. Where, recognising its vulnerability as a global financial hub, it has adopted a defense-centric deterrence strategy by focusing on redundancy, cyber education, and international partnership rather than offensive capacity. These approaches can also showcase that deterrence is not always about scaring attackers with retaliation, it is about making the attacks meaningless.  

‍

Cyber deterrence and Asymmetric Warfare

‍

Cyber conflict is understood through the lens of asymmetric warfare, where weaker actors exploit the unconventional and stronger foes. As guerrillas get outmanoeuvred by superpowers in Vietnam or Afghanistan, small states hold the capability to frustrate the cyber giants by turning their size into a shield. The essence of asymmetric cyber defence also lies in three principles, which can be mentioned as; 

• Resilience over retaliation by ensuring a rapid recovery to neutralise the goals of the attackers. 
• Undertaking smart investments focusing on limited budgets over critical assets, not sprawling infrastructures. 
• Leveraging norms to shape the international opinions to stigmatize the aggressors and increase the reputational costs. 

This also helps to transform the levels of cyber deterrence into a game of endurance rather than escalating it into a domain where small states can excel. 

There remain challenges as well, as attribution problems persist, the smaller nations still depend on foreign technology, which the adversaries have sought to exploit. Issues over the shortage of talent have plagued the small states, as cyber professionals have migrated to get lucrative jobs abroad. Moreover, building deterrence capability through norms requires active multilateral cooperation, which may not be possible for all small nations to sustain. 

Conclusion

‍

Cyberwarfare represents a new frontier of asymmetric conflict where size does not guarantee safety or supremacy. Great powers have often dominated the offensive cyber arsenals, where small states have carved their own path towards security by focusing on defence, resilience, and international collaboration. The examples of Singapore and Estonia demonstrate the fact that the small size of a state can be its identity of a hidden strength in capabilities like cyberspace, allowing nimbleness, concentration of resources and societal cohesion. In the long run, cyber deterrence for small states will not rest on fearsome retaliation but on making attacks futile and recovery inevitable. 

‍

References

‍

• https://bluegoatcyber.com/blog/asymmetric-warfare/
• https://digitalcommons.usf.edu/cgi/viewcontent.cgi?article=2268&context=jss
• https://www.linkedin.com/pulse/rising-tide-cyberwarfare-battle-between-superpowers-hussain/ 
• https://digitalcommons.odu.edu/cgi/viewcontent.cgi?article=1243&context=gpis_etds
• https://www.scirp.org/journal/paperinformation?paperid=141708
• https://digitalcommons.odu.edu/cgi/viewcontent.cgi?article=1243&context=gpis_etds