Are DeepFakes Illegal?: Legal Challenges, Global Efforts, and Detection Challenges

What started in 2017 as the name of a subreddit on popular social media platform “reddit” for creation and sharing of face-swapped pornography has since evolved into the widely understood and recognized term for all altered media by the help of artificial intelligence tools. The deletion and end of the subreddit did not help in stopping the term “deepfake” from gaining traction. Although the history of deepfakes is much more complicated to navigate than the subreddit, the popularity of the term in modern times can be understood from this context. 

 A deepfake is a video, image or audio which was fabricated with the help of AI powered tools to represent itself as something it is actually not. This includes but is not limited to face swapping, fake narratives and revenge porn. The use case of a deepfake is statistically more malicious than beneficial as it is seen repeatedly being used to present reality as something which it is not, to spread lies or confusion among its target audience. 

Recent years have seen a surge in deep-fakes and AI generated images across the globe, whether it is a celebrity face swap, revenge deep-fake porn, corporate scam, or a media generated during elections with malafide intent to defame and derail the target’s political campaign, deep-fakes know no boundaries or exceptions. It is an unhinged and often weaponized technology used with unreasonable intentions and the ease of access, simple user interface(s) alongwith rapid improvements in the quality of fake media are not helping mitigate the enormous dangers it poses.

A recent alert issued by USA’s FINCEN ( Financial Crimes Enforcement Network) dated Nov 13, 2024 highlights the increasing use of deepfake media in fraudulent activities targeting financial institutions. The alert emphasized on how criminals are leveraging deepfakes to create falsified documents, photographs and videos to bypass customer identification and verification processes. This alert raised the urgent need for vigilance in detection and prompt reporting of such activities to protect the financial system.

Similarly, Australia’s Scam Prevention Framework which was enacted in February 2025, expanded on the dangerous threat of scams using Artificial Intelligence and proposed imposing of up to $50 million on social media companies that fail to remove scams involving AI or deepfake technology. 

1. Impact of DeepFake Attacks:

A deepfake attack is an orchestrated targeting fueled by malicious intent. Deepfake attacks have seen a significant rise in the past few years. Deepfake technology can be used for a wide variety of appalling purposes and recent years have seen a surge in such crimes impacting common citizens to corporations alike. 

In a shocking case from Spain in 2024, a viral video showed a distressed teenage girl begging for help, only for investigators to find out later that she was entirely AI generated, a hoax designed to manipulate masses.

Similarly, in 2023 in India, around 100 actresses and influencers fell victim when their faces were swapped into pornographic videos, which spread instantly to all corners of social media. The outrage forced India’s hand for stricter laws, but nevertheless at the cost of irreparable damage to victims’ reputations.

Elections are no exception either. In the New Hampshire 2024 elections, voters received robocalls mimicking President Biden, falsely telling them to skip the voting, marking one of the first known cases of AI powered election interference in the US. Meanwhile, criminals are using AI to clone voices of loved ones or bosses, like the 2019 scam of $243 million against a German energy firm where fraudsters impersonated the CEO to trick an employee into transferring funds. 

Similarly, criminals now use deepfakes to spoof facial recognition systems of banks and financial institutions for stealing identities and draining accounts with exceptional ease.

These types of attacks are not just a mere categorization of the possibilities now but something which is already seen to happen repeatedly over the years and requires extensive overseeing and effective countermeasures. 

Deepfakes pose a significant threat to reality because they present a person or a situation as it is actually not. While the cons of the technology outweigh pros, some positive use cases might still be argued in this regard, such as; use in the entertainment industry for high quality CGI, voiceovers, video game characters without costly reshoots, or for the purposes of education, marketing, accessibility and art. 

2. Are Deepfakes legal? Global Regulations and Challenges

The legality of deepfakes and the regulations surrounding it are continuously being developed and enforced around the globe. Jurisdiction, use case, purpose and intention of creating a deepfake are some of the factors which contribute to the legality of deepfakes. While not explicitly illegal around the globe, we are seeing significant regulatory efforts to better define illegality of deepfakes and ways to combat them. Some significant legislation efforts around the globe are:

United Kingdom: 

– Data Use & Access Bill (2025) : Criminalizes creating sexually explicit deepfakes without consent, with offenders facing unlimited fines.  

– Online Safety Act (2025) : Treats non-consensual deepfakes as a “priority offense”, forcing platforms to proactively remove them. 

United States (Federal):

– Take It Down Act (Pending): Criminalises sharing non-consensual explicit deepfakes, requiring removal within 48 hours of reporting

– No Fakes Act (Proposed): Aims to protect against unauthorized digital replicas, setting federal standards for AI-generated impersonations. 

– Defiance Act (Proposed 2024): Gives victims legal power to sue creators and distributors of harmful deepfake pornography. 

– AI Labeling Act (Proposed 2023): Would require clear labeling of AI-generated content to combat misinformation, 

United States (State level): 

– California: Allows lawsuits for non-consensual explicit deepfakes (AB 602) and bans election-related deepfakes (AB 730). New laws (SB 942, AB 2655) demand AI watermarking and combat political deepfake deception. 

– Tennessee ( ELVIS Act 2024 ): Protects artists by banning unauthorised AI voice/likeness cloning.

 European Union:

– AI Act (2024): Forces clear labeling of AI generated content, including deepfakes. 

– Digital Services Act (2024): Requires platforms to tackle illegal deepfakes proactively. 

France: 

– SREN Law (2024): Explicitly bans non-consensual deepfakes, punishing distribution under criminal law.

Canada: 

– Intimate Images Protection Act: Targets distributors of non-consensual deepfake pornography but does not penalize creators.

UAE: 

– AI Ethics Charter: Promotes transparency and accountability in AI, discouraging malicious deepfake use. 

Spain:

– AI Transparency Law (Proposed): Would impose heavy fines of up to €35M for unlabeled AI content and ban manipulative deepfakes.

China: 

– Deep Synthesis Regulations (2023): Mandates watermarking AI content ensuring users know when media is synthetic.  

While deepfake laws vary globally but common themes such as;  consent, transparency, intent and accountability emerge repeatedly. Many nations have now criminalized harmful deepfakes especially in revenge porn and elections while pushing for AI watermarking and platform responsibility. However enforcement of these laws still remains a challenge as the technology involved in creation significantly outpaces the legislative efforts. 

3. Detection Efforts & Regulations:

The rise of deepfakes has sparked a worldwide push to develop better detection tools and governments, tech companies and researchers are stepping up. From AI-powered detection tools to public awareness campaigns. 

Different nations are deploying different tackling measures to the challenge of detecting deepfakes to better protect against deepfake attacks.

In the UK, regulators are pushing tech companies to take prompt action. Ofcom’s draft Online Safety Guidance (2025) recommends that platforms use “hash-matching technology” , a digital fingerprinting system to flag and block deepfakes before they go viral. Meanwhile, Norway is taking a different approach i.e instead of just relying on tech, they have launched nationwide deepfake awareness campaigns teaching people how to scrutinize suspicious videos through media literacy programs.

Over in the Netherlands, a specialized “Deepfake Detection Task Force” is working on both “real-time and forensic tools” to catch manipulated media especially when it threatens elections or public trust. Similarly the US has been investing in AI forensics for years, programs like DARPA’s “MediFor” laid the groundwork for analyzing digital tampering, while newer initiatives like the “AI FORCE” are crowdsourcing cutting-edge detection tools and algorithms. 

Moreover, the EU can also be seen evidently joining the fight. EU’s “Code of Practice on Disinformation” pushes platforms to prioritize authentic content while funding research into better detection methods such as analysing facial glitches or inconsistent audio.

Behind the scenes, scientists at NIST US are developing frameworks to spot AI generated images and content by digging into “metadata and digital artifacts”. 

4. Conclusion:

As seen from examples and references the legality of deepfakes around the globe is still a developing and improving area. This gray area highlights the importance of efforts at individual as well as corporation level to implement and use effective cybersecurity measures to detect and prevent deepfake misinformation or scams. Deepfake attacks for the purpose of scamming and defaming have risen significantly whereas the rapidly improving technology and ease of access is not helping the cause of the fight against these attacks. 

The legislative efforts and proposed bills around the globe will likely make watermarking and labeling mandatory in the near future. Meanwhile, increasing one’s knowledge on the difference between a real and deepfake media along with the use of an accurate and cutting-edge tool such as the one provided by FACIA will help greatly in providing a fighting chance against hyper-realistic fakes because in the deepfake era, seeing does not always mean believing.