Deepfakes use AI to manipulate media including audio or video to create hyper-realistic, but entirely fake, content of

real people. While they can be entertaining or even harmless in some cases, they can also pose real threats—including
political manipulation, fraud and invasion of privacy.

The advancement of artificial intelligence and deepfake technology has made it easier than ever to create convincing
false content. And with misinformation and disinformation on the rise in our politically polarized modern society, the
consequences can be dangerous. Whether it’s a celebrity endorsing a cause they don’t support or someone tricking a
financial institution with a fake voice, deepfakes are becoming a serious issue in today’s digital world.

Political Deepfakes

In July of 2024, Elon Musk, CEO of Tesla and owner of X (formerly Twitter), shared to his 199.2 million X followers a
deepfake video of vice president and presidential candidate Kamala Harris cursing and saying out of character
statements. One month later, former president and president elect Donald Trump shared to his 7.78 million followers on
his Truth Social platform a deepfake image of popular singer Taylor Swift endorsing his 2024 presidential campaign,
along with several images of digitally rendered young women in “Swifties for Trump” t-shirts. These deepfakes quickly
went viral.

Aside from being dishonest, deepfakes depicting public figures can polarize public opinion, create chaos and even
influence elections—and at lightspeed! A deepfake can go from a few retweets to global confusion in a matter of hours.

While some view this type of content as just funny and harmless parody, when being shared by public and political
figures with a large reach and depicting public and political figures with an outsize impact on our elections (when Taylor
Swift did eventually endorse a candidate, 35,000 new voters registered within 24 hours), the potential for confused
voters is quite high.

No matter where you fall on the political spectrum, deepfakes are an unfortunate obstacle in the pursuit of truth.

Voice Deepfakes for Fraud

It’s not just video content that’s used in AI manipulation. Voice deepfakes have emerged as a new and troubling trend in
scams. One high-profile case involved a UK-based energy firm being scammed out of $243,000 when criminals enacted a
“vishing” campaign (“vishing” is short for “voice phishing,” the tactic of tricking targets over the phone) by using an
artificial voice so similar to the boss of the German parent company of the firm that the employees didn’t notice a
difference. This incident marks the first time AI-based voice fraud has netted such a high payload, according to The Next
Web.

By mimicking a familiar voice, scammers were able to execute a financial heist with shocking precision. This kind of
attack highlights the growing threat to both financial and personal security. When you can’t trust the voices you hear,
verifying identities becomes a major challenge for businesses and individuals alike.

The Crisis of Deepfake Pornography

The dark side of deepfake technology also invades people’s most personal spaces. Deepfake pornography, where a
person’s face is superimposed onto explicit content without their consent, has become a growing issue,
disproportionately affecting women. Politician Alexandria Ocasio-Cortez has been the target of deepfake pornography
and one recent study found that 98 percent of deepfake videos online were pornographic and that 99 percent of those
targeted were women or girls.

This misuse of AI is a serious violation of privacy and personal safety. It not only tarnishes reputations but also causes
psychological harm. Victims are often left with little legal recourse (there are many states that have not criminalized
deepfake revenge porn), and the damage done to their personal and professional lives can be irreversible.

Deepfake Ads and False Endorsements

Deepfakes are also now creeping into the world of advertising. These AI-generated ads can mislead consumers into
thinking their favorite celebrity is behind a product or service when, in fact, they aren’t involved at all.

These deepfake ads present major legal and ethical concerns. They can tarnish a celebrity’s brand and lead to lawsuits
over false advertising. For consumers, it’s yet another layer of deception to navigate in an already complex digital
advertising world.

5 Practical Tips for Spotting Deepfake Content

So how can you protect yourself from falling for a deepfake? Here are a few practical tips:

1. Look for unnatural eye movements – Deepfake technology struggles to replicate the natural blink rate and
subtle eye movements of real humans.

2. Check for inconsistencies in lighting – If the lighting on the person doesn’t match the background, it might be a
deepfake.

3. Watch for odd facial expressions – Deepfakes often have facial movements that seem slightly off, like a smile
that doesn’t quite match the tone of the conversation.

4. Listen closely – If it’s a voice deepfake, the tone may be flat, or the speech pattern might sound unnatural.

5. Cross-reference sources – If a piece of content looks suspicious, check it against trusted news sources to confirm
its authenticity.

A Future with Deepfakes

Deepfakes are no longer just the stuff of science fiction – they are here, and they’re causing real harm. From political
manipulation and financial fraud to privacy violations and misleading ads, the potential for misuse is vast. As we’ve seen
from these examples, the key lesson is to remain vigilant and critical of the content we consume—always investigate
sources and double-check with other news outlets if an image or audio seems suspect.

The future of deepfakes will continue to challenge our legal systems, ethical standards, election processes and even the
way we perceive reality. As the technology behind deepfakes evolves, so must our strategies for detecting and
preventing their misuse.

Detect DeepFakes: How to counteract misinformation created by AI

Detect DeepFakes: How to counteract misinformation created by AI was active from April 2020 to January 2025

How do you spot a DeepFake? How good are DeepFake videos? How well can ordinary people tell the difference
between a video manipulated by AI and a normal, non-altered video? Rather than try to explain in words, we built
the Detect Fakes website so you can see the answer for yourself. Detect Fakes is a research project designed to answer
these questions and identify techniques to counteract AI-generated misinformation. It turns out there are many subtle
signs that a video has been algorithmically manipulated. Some subtleties are explained in detail below.

We already know DeepFakes can be quite believable, but just how believable are they? Kaggle's Deepfake Detection
Challenge (DFDC) recently sought an algorithmic answer to this question of detecting fakes. The description on the
Kaggle Website explains, "AWS, Facebook, Microsoft, the Partnership on AI’s Media Integrity Steering Committee, and
academics have come together to build the Deepfake Detection Challenge (DFDC). The goal of the challenge is to spur
researchers around the world to build innovative new technologies that can help detect deepfakes and manipulated
media." The winners of the Kaggle Competition were awarded $1,000,000.

Rather than fine-tune the best machine learning model for this Kaggle competition, we are curious about strategies and
techniques for building public awareness of DeepFake technology and helping ordinary people think critically about the
media that they consume.
We hypothesized that the exposure of how DeepFakes look and the experience of detecting subtle computational
manipulations will increase people's ability to discern a wide-range of video manipulations in the future. As such, we
hosted a website called Detect Fakes to display thousands of the curated, high-quality DeepFake and real videos from
the DFDC dataset publicly. The latest version of the website shows 32 videos that were produced as part of the
Presidential Deepfakes Dataset.

The Detect Fakes experiment offers the opportunity to learn more about DeepFakes and see how well you can discern
real from fake. When it comes to AI-manipulated media, there's no single tell-tale sign of how to spot a fake.
Nonetheless, there are several DeepFake artifacts that you can be on the look out for.

1. Pay attention to the face. High-end DeepFake manipulations are almost always facial transformations.

2. Pay attention to the cheeks and forehead. Does the skin appear too smooth or too wrinkly? Is the agedness of
the skin similar to the agedness of the hair and eyes? DeepFakes may be incongruent on some dimensions.

3. Pay attention to the eyes and eyebrows. Do shadows appear in places that you would expect? DeepFakes may
fail to fully represent the natural physics of a scene.

4. Pay attention to the glasses. Is there any glare? Is there too much glare? Does the angle of the glare change
when the person moves? Once again, DeepFakes may fail to fully represent the natural physics of lighting.

5. Pay attention to the facial hair or lack thereof. Does this facial hair look real? DeepFakes might add or remove a
mustache, sideburns, or beard. But, DeepFakes may fail to make facial hair transformations fully natural.

6. Pay attention to facial moles. Does the mole look real?

7. Pay attention to blinking. Does the person blink enough or too much?

8. Pay attention to the lip movements. Some deepfakes are based on lip syncing. Do the lip movements look
natural?

These eight questions are intended to help guide people looking through DeepFakes. High-quality DeepFakes are not
easy to discern, but with practice, people can build intuition for identifying what is fake and what is real. You can
practice trying to detect DeepFakes at Detect Fakes.

Dangers of Deepfake: What to Watch For

In the past few years, artificial intelligence technology has crossed a threshold with the capability to make people look
and sound like other people. A “deepfake” is fabricated hyper-realistic digital media, including video, image, and audio
content. Not only has this technology created confusion, skepticism, and the spread of misinformation, deepfakes also
pose a threat to privacy and security.

With the ability to convincingly impersonate anyone, cybercriminals can orchestrate phishing scams or identity theft
operations with alarming precision. In a recent incident, cybercriminals posed as a company’s chief financial officer and
other colleagues in a Zoom meeting. The elaborate scam led to the loss of $25 million.

Identify and report deepfakes

To help protect Stanford against similar attempts, below are some tips to identify and report deepfakes.

What you should watch for:

 Be wary of phone calls or videos that appear to come from trusted colleagues or senior executives when they
involve unexpected demands or requests for financial transactions.

 Take note of unusual, urgent requests to set up vendors, make purchases, process electronic fund transfers, or
update accounts payable or banking information.
What you should do:

 If you have any doubts about the legitimacy of a request, pause and consult with others before proceeding. It’s
always better to take a moment to verify than to fall victim to a fraudulent scheme.

o Independently verify the request by contacting the alleged requester using a separate, trusted
communication channel that you initiate. This will ensure that you’re communicating with a legitimate
person, rather than an imposter.

 If you encounter any suspected deepfake fraud attempts or fraudulent activity, immediately report it. Here are a
few Stanford resources for reporting:

o Information Security Office and University Privacy Office for IT security or privacy concerns (e.g.,
suspected phishing scam or data compromise).

o Ethics and Compliance Hotline for concerns about possible financial irregularities, unethical behavior,
and violations of policies, regulations, or laws.

o Fingate (view under Managing Potential Fraud and Disputes) if you suspect the deepfake to be initiated
by someone who has hacked one of our suppliers.

Share this information with members of your team or business unit who initiate and approve financial transactions. Your
vigilance and communication will help protect our community against the threat of deepfake fraud and maintain the
integrity of our financial transactions.

You've probably heard about generative AI (artificial intelligence), but you might not be aware that these technologies
have brought with them new concerns about privacy, identity theft, and misinformation. One pernicious form of AI scam
is called "deepfake" technology.

Deepfakes are artificial intelligence-generated videos or audio clips that make it appear as though someone is saying or
doing something they never did. Just by this definition, the possibilities for identity theft and misinformation might
become obvious to you. Deepfakes can be used to defame individuals and commit fraud. For example, if your vocal
identity and sensitive information got into the wrong hands, a cybercriminal could use deepfaked audio to contact your
bank.

You might think that because you don't use any AI product you could never be a victim. The truth is that these
technologies can scrub data (such as video, photographs, and voice recordings) of millions of people from websites, like
social media platforms.

You can take some steps to reduce the chances that a criminal creates a deepfake of you. Mostly, you should think hard
about what you share publicly. Here are some strategies to protect yourself, and some tips about what to do if you
suspect you're the victim of a deepfake.

1. Share with care: The first step in avoiding deepfakes is to be extremely cautious about what personal
information you share online. Limit the amount of data available about yourself, especially high-quality photos
and videos, that could be used to create a deepfake. You can adjust the settings of social media platforms so
that only trusted people can see what you share. Of course, you should also make sure that you trust anyone
who requests to follow or friend you.

2. Enable strong privacy settings: Take full advantage of websites' privacy settings to control who can access your
personal information and content. Restrict who can see your photos, videos, and other sensitive data. This
includes websites where you store photo files. Reduce the amount of publicly available material, and you
minimize the resources potential deepfake creators have.
 3. Watermark photos: When sharing images or videos online, consider using a digital watermark on them. This can
discourage deepfake creators from using your content since it makes their efforts more traceable.

4. Learn about deepfakes and AI: The realm of AI is changing rapidly. Staying abreast of the latest developments
can help you stay vigilant. You don't need to become an expert, but following the news about these
technologies is important for everybody. This knowledge can help you recognize potential red flags when
encountering suspicious content.

5. Use multi-factor authentication: These days, you really should double your security by implementing multi-
factor authentication for all of your accounts. This is when you need an extra step to log into an account, such as
a facial scan, entering a code texted to your phone, or using a standalone app on your device. This extra layer of
security helps prevent unauthorized access to your accounts, reducing the chances of someone obtaining your
personal data.

6. Use long, strong, and unique passwords: Every password should be at least 16 characters long, unique to the
account, and contain a random mix of upper case letters, lower case letters, numbers, and special characters.
The best way to remember all these unique passwords is by storing them in a password manager with MFA
turned on.

7. Keep your software up to date: Keep your devices and software up to date with the latest security patches and
updates. Outdated software can have vulnerabilities that hackers may exploit to access your data. We
recommend turning on automatic updates so you don't have to constantly check for new updates.

8. Don't take the phishing bait: Be extremely cautious when receiving emails, direct messages, texts, phone calls,
or other digital communications if the source is unknown. This is especially true if the message is demanding
that you act fast, such as claiming your computer has been hacked or that you won a prize. Deepfake creators
attempt to manipulate your emotions so you download malware or share personal information. Verify the
identity of the sender and avoid clicking on suspicious links. We always say: think before you click.

9. Report deepfake content: If you come across deepfake content that involves you or someone you know, report
it to the platform hosting the content. This can help in having it removed or investigated, limiting its potential
reach. You should also report it to federal law enforcement.

10. Consult legal advice: If you are the victim of a deepfake that has damaged your reputation, consult with
cybersecurity and data privacy legal experts. Laws are quickly evolving to address the issue of deepfakes, but
with technology changing so fast, the legal system takes time to catch up. You can take action – contact your
elected representatives and tell them you want to see more action around preventing deepfakes!

While deepfakes provide a new terrain in the battle against misinformation and defamation, you can take proactive
measures to protect your digital identity. In fact, this advice is good for protecting yourself from other common cyber
threats. Stay informed, adopt these cybersecurity habits, and think hard about what you post online, as well as who has
access to it.

Imagine being able to see a video of your favorite actor saying something they never said, or a politician admitting to a
crime they never committed. Fascinating and terrifying at the same time, right? Welcome to the intriguing world of
deepfakes, a technology that is blurring the lines between reality and fiction. On this digital labyrinth journey, we'll
unravel the mystery behind deepfakes: how a mix of artificial intelligence and digital creativity is creating a new era of
audiovisual content.

Deepfakes, deriving their name from the combination of deep learning and fake, are more than mere audiovisual
manipulations; they are a testament to the exponential advancement of technology and a reminder of the fragility of
our perception.
Throughout this post, you'll not only discover how these impressive digital works are created, but also why it's essential
to develop a critical mindset and an interrogative attitude in this digital information era.

What is a deepfake?

Deepfakes are an advanced form of digital manipulation using artificial intelligence and machine learning to create or
alter audiovisual content. At their core, a deepfake can make it seem as if someone has said or done something that
never actually happened, producing videos or audios that are incredibly difficult to distinguish from the real ones.

This technology, though it may seem like something out of a science fiction movie, is surprisingly accessible and has
gained popularity in recent years. Deepfakes have been used in a variety of contexts, from meme creation to
manipulation in political environments, to advertising, like the award-winning Cruzcampo (Spanish beer) commercial
with Lola Flores, underscoring the need for critical analysis and a detailed understanding of their nature.

The origin of deepfakes

The term deepfake originated around 2017, when an internet forum user with the alias deepfake began posting
manipulated videos using artificial intelligence software. These early deepfakes often superimposed celebrity faces onto
other people's bodies in videos, quickly generating a mix of awe and concern over their ethical and legal implications.

Since then, the technology has evolved rapidly. What began as a novelty in online forums has transformed into a tool
with the potential to create significant changes in the way we consume and trust audiovisual content.

Deepfakes and artificial intelligence

Creating a deepfake is a complex process that relies on the use of artificial intelligence algorithms, specifically those
focused on deep learning. These algorithms analyze thousands of images and videos to learn how to mimic a person's
facial expressions, movements, and voice. This information is then used to synthesize new images or video sequences
that appear real but are completely fabricated.

This ability of AI to replicate and alter reality raises significant questions about authenticity and trust in digital media.
While deepfakes can have creative and entertaining applications, they also represent a serious challenge in terms of
security, privacy, and ethics in the digital age.

How are deepfakes created?

Creating a deepfake is a fascinating process that combines advanced technology and creative skills. At first glance, it
might seem like digital magic, but behind each deepfake is a detailed set of technical steps and artistic decisions.

Deepfake creation process

The process begins with the collection of a large amount of data, specifically images and videos of the person to be
replicated. The more material collected, the more realistic and convincing the final result will be. This data is fed into an
artificial intelligence system, which uses deep learning techniques to analyze and understand the person's facial
features, movements, and expressions.

Then, a second dataset is used, which usually consists of the material in which the person's image or voice is to be
inserted. The AI system, through complex algorithms, merges these two datasets, creating a representation that looks
natural and realistic.

This fusion is not a straightforward process and requires meticulous adjustments. The synchronization of lip
movements, facial expressions, and lighting are essential for the deepfake to be convincing. Additionally, it's necessary
to adjust the audio, especially if creating a video where the person appears to be speaking.

Technology behind deepfakes

The key technology behind deepfakes is deep learning, a branch of artificial intelligence. Specifically, convolutional
neural networks (CNN) and generative adversarial networks (GAN) are used to process and generate the images and
videos.

CNNs are effective in recognizing and replicating patterns in images, such as facial features. GANs, on the other hand,
are a set of two neural networks that work together: one generates the image, while the other evaluates its
authenticity, in a continuous improvement process until achieving a result that is difficult to distinguish from reality.

This sophisticated use of AI not only demonstrates the power of modern technology but also highlights the need for
deep understanding and careful regulation, given the potential for misuse of these tools.

The impact of deepfakes on society

Deepfakes are not just a technological feat; their impact extends far beyond the digital realm, affecting fundamental
aspects of our society. These digital creations have the potential to influence public opinion, alter the perception of
reality, and challenge the authenticity of information.

Deepfakes in politics and media

In the political and media arena, deepfakes represent a powerful tool for disinformation and manipulation. A notable
case was the deepfake of Ukrainian President Volodymyr Zelensky, where he supposedly asked Ukrainian forces to
surrender to the Russian army. Discussions have also centered on deepfakes related to Russian President Vladimir Putin,
highlighting the sophistication and political manipulation potential of this technology.

Identity impersonation with deepfakes

Beyond politics, deepfakes pose a serious risk of identity impersonation. In entertainment, the TikTok account
"deeptomcruise" created deepfakes of actor Tom Cruise, deceiving millions. In the business world, a deepfake was used
to impersonate the head of communications at Binance, aiming to deceive representatives of other companies in
cryptocurrency transactions. These cases demonstrate the threat deepfakes pose in terms of security and privacy.

How to Detect and Protect Yourself from Deepfakes

Knowing how to identify and protect yourself from deepfakes is crucial, especially as the technology advances and
becomes even more sophisticated. Here are some key tips for spotting these forgeries:

 Inconsistencies in audio and video: Pay attention to mismatches in lip movement or facial expressions that do
not match the audio.

 Quality and resolution: Deepfakes often have flaws in image quality or resolution, particularly around the eyes
and mouth.

 Lighting and shadows: Check if the shadows and lighting on the person's face do not align with the
environment.

 Blinking and natural movements: People in deepfakes may not blink naturally or have unusual facial
movements.

Prevention against the Threat of Deepfakes

To prevent the impact of deepfakes, it's essential to adopt a critical and questioning stance towards the information we
consume:

 Source verification: Always verify the source of a video, especially if the content is surprising or provocative.

 Education and awareness: Stay informed about deepfake technologies and teach others to be critical of the
content they consume.
Tools for Deepfake Detection

Various tools and software have been developed to help identify deepfakes. Here are some of the most well-known:

 Sentinel: Learn about the AI-based protection platform Sentinel and its use in detecting deepfakes.

 FakeCatcher by Intel: Discover how Intel's FakeCatcher detects deepfakes with high precision.

 WeVerify: Explore WeVerify for misinformation analysis and content verification.

 Microsoft Video Authenticator: Learn about Microsoft's Video Authenticator for identifying deepfakes.

 TensorFlow and PyTorch: Use TensorFlow and PyTorch in deepfake detection.

Protect Your Digital Identity from Deepfakes with Didit

In the fight against deepfakes and the protection of digital identity, tools like Didit from Gamium become fundamental
allies. This decentralized digital identity solution offers several layers of security and verification that can help safeguard
the authenticity and integrity of personal and business identity in the digital environment.

Key Features of Didit in Protecting Against Deepfakes

 Robust authentication: Didit provides a strong authentication system, ensuring that only the true owner of the
digital identity can access and use it.

 Privacy control: With Didit, you have total control over your personal data, meaning you can manage who has
access to your information and to what extent.

 Transparency and security: By employing decentralized technology and the industry's highest encryption, Didit
ensures data transparency and security, which is crucial in preventing forgeries and manipulations.

How Can Didit Help in Practice?

1. Prevention of identity impersonation: By securing your decentralized identity, Didit can prevent impostors from
creating deepfakes that impersonate you online.

2. Verification on social media and online platforms: Didit can be used to verify the authenticity of profiles and
content on social media and other platforms, essential for avoiding being deceived by deepfakes.

3. Security in business transactions: In the business world, Didit ensures that transactions and communications
are authentic, protecting businesses from deepfake-based frauds.

Create Your Digital Identity with Didit

Integrating Didit into your digital security strategy not only protects your online identity, but also provides an additional
layer of defense against the sophisticated threats posed by deepfakes. In a world where the line between reality and
fiction is increasingly blurred, tools like Didit are indispensable for maintaining integrity and trust in the digital space.

And you're just one step away from better protecting yourself from deepfakes. Click the button below and create your
decentralized digital identity. Only you will have the power over your private information.