Please ensure Javascript is enabled for purposes of website accessibility

Deepfake Blog

Banks vs Deepfakes: The Financial Sector Under Pressure

Imagine receiving a video call from your bank’s representative, their face and voice completely familiar, requesting urgent authorization for a large transaction. Some people may be overcautious and run checks, but others may fall for the scam. We think the risks are substantial (and growing).

Allon Oded, VP Product
March 10, 2024

Guarding the C-Suite: Deepfakes and Brand Risk

Imagine a video circulating online, featuring the CEO of a major financial institution endorsing a dubious investment scheme. The video looks and sounds authentic, but it's a meticulously crafted deepfake. This scenario, once the stuff of science fiction, is becoming an increasingly possible reality, posing a significant threat to brand reputation and financial security.

Michael Matias, CEO
January 15, 2025

Deepfakes Enter the Battlefield: Propaganda in Conflict

As digital manipulation blurs the lines between reality and fiction, deepfake technology has emerged as a formidable weapon in the arsenal of information warfare. Malevolent actors increasingly rely on deepfakes to convince and compel their targets.

Niv Amitay, AI Generation Team Lead
December 5, 2023

Regulators Tackle Deepfakes: New Laws and Policies

Regulatory advancement can combat deepfake use. This article examines the key regulatory developments of 2023, highlighting the urgent responses and emerging trends that shaped the landscape.

Michael Matias, CEO
January 10, 2024

AI-Powered Deepfake Detection Advances

The spread of sophisticated deepfakes presents a critical threat: fortunately, advancements in deepfake protection technology are providing effective solutions, offering a vital shield against these deceptive practices.

Ilia Kravets, VP Engineering
August 20, 2023

Next-Gen Deepfakes And Insurance Coverage

The deepfake threat is significant, and the consequences can be devastating and require insurance. As a Reuters report highlights, the need for cyber insurance matters – but insurance will only cover organizations to a limited extent, actively combatting deepfakes is critical too.

Allon Oded, VP Product
October 14, 2024

Incorporating Deepfakes into Crisis Management Plans

As cybersecurity measures struggle to keep up with the rise of deepfakes there’s a growing probability of attack success. Executives should consider proactively integrating deepfake scenarios into their crisis management plans to ensure they’re able to respond to an attack and mitigate the consequences.

Michael Matias, CEO
July 2, 2024

Australia Outlaws AI Deepfake Porn in Online Safety Reforms

Australia just added to the global momentum to regulate harmful deepfakes, setting a legal precedent in the Asia-Pacific region. New legislation imposes a duty to respond quickly to reports of deepfake porn.

Doron Ish-Shalom, Head of Strategic Partnerships
June 12, 2024
NSA Launches AI Security Center to Shield U.S. from AI-Related Threats

NSA Launches AI Security Center to Shield U.S. from AI-Related Threats

As AI steadily becomes core infrastructure, protecting it is as mission-critical as securing networks or databases. The NSA’s AI Security Centre validates this and is set to help ensure AI systems can be trusted and safeguarded against emerging threats.

Lior Ben Moha, Data Ops Engineer
September 29, 2023

Thousands of Hackers Red-Team Generative AI Models at DEF CON 31

At the recent 2023 DEF CON hacker convention the cybersecurity community stress-tested LLMs, taking a proactive approach to uncovering AI vulnerabilities (from jailbreaking LLMs to prompt injection attacks) before malicious actors do.

Yinon Douchan, AI Researcher
August 29, 2023

UK Online Safety Act Criminalizes Non-Consensual Deepfake Pornography

Turning harmful actions into a criminal offence is sometimes the only way to deter bad actors. The UK government now took concrete action to tackle a malicious use of deepfakes that technical measures alone haven’t solved.

Shir Koren, Software Engineer
October 26, 2023

US FEC Moves to Potentially Regulate AI-Deepfakes in Political Ads Ahead of 2024 Election

When election regulators sound the alarm on deepfakes, it underscores the extent of misuse of AI. Clarity believes transparency in AI-generated political content isn’t just ethical – it’s table stakes for preserving public trust in the electoral process.

Michael Matias, CEO
March 28, 2024

White House Secures Voluntary Pledges from AI Firms on Deepfake Safeguards

We know we need guardrails around AI use – but can companies voluntarily agree to apply controls to reduce the risk of malevolent AI use? The US Government appears to have made some progress in achieving that.

Doron Ish-Shalom. Head of Strategic Partnerships
September 14, 2024

Los Alamos Researchers Break Record in AI Malware Detection

Catching malware can be like finding a needle in a haystack. This report from LANL shows how AI can shrink that haystack – a win for companies trying to guard against a growing cybersecurity threat.

Eli Passov, AI Tech Lead
February 15, 2024

Meta Develops Voicebox AI That Mimics Voice – But Withholds It Over Misuse Risks

Is it worth pausing the broader release of an AI tool until robust safeguards and ethical guidelines are developed? It’s an important question as AI continues to advance – the potential for misuse is simply too large.

Talia Ben Simon, AI Researcher
June 22, 2023

FBI Warns of Surge in AI Scams

The FBI’s warning confirms what we’ve been seeing: AI is turbo-charging fraud at scale – a clear signal for businesses to strengthen multi-factor verification and stay vigilant against synthetic media scams.

Nir Gofman, Full Stack Engineer
December 3, 2024

“WormGPT” Emerges for Phishing Attacks

The emergence of WormGPT proves that cybercriminals are embracing AI to scale attacks – that’s why, at Clarity, we are innovating as fast as we can with AI-driven detection that keeps pace with an evolving cybersecurity threat.

Michael Matias, CEO
September 19, 2023

$25M Deepfake CEO Scam Shakes Hong Kong Firm

Digital trust is increasingly fragile, a recent Hong Kong deepfake scam serves as a reminder of the need for robust AI-powered detection and verification tools that can help organizations guard against deepfake attacks.

Doron Ish-Shalom, Head of Strategic Partnerships
May 20, 2024

White House Executive Order Demands Watermarking of AI Content

At Clarity, we continue to work to a future where AI does no harm, but it is a shared responsibility. The recent surge in deepfakes underscores the critical need for government mandates around AI solutions, such as watermarking, that can combat misinformation and ensure content authenticity.

Lior Ben Moha, Data Ops Engineer
December 4, 2023

Google Unveils SynthID for Watermarking AI Images

Developing effective methods to identify AI-generated content represents an important step in ensuring responsible AI use. At Clarity, we monitor advances in content authentication technologies as part of our commitment to addressing synthetic media challenges. Google DeepMind's launch of SynthID, a watermarking tool that identifies AI-generated images, represents a development in this field. By providing a mechanism to identify AI content, tools like SynthID can help reduce potential misuse while maintaining the benefits of generative technologies.

Nir Gofman, Full Stack Engineer
August 30, 2023

DARPA Announces AI Cyber Challenge at Black Hat

Artificial intelligence offers potential for addressing complex cybersecurity challenges, but effective implementation requires coordinated effort across the industry. At Clarity, we recognize the importance of collaborative innovation in this space, which is why we're encouraged by DARPA's investment in the AI Cyber Challenge. This initiative represents an important step toward developing new approaches to vulnerability detection and remediation through competitive innovation..

Lioe Ben Moha, Data Ops Engineer
August 10, 2023

Google's Sec-PaLM: How the New AI Workbench Transforms Enterprise Cybersecurity Operations

At the RSA Conference 2023, Google unveiled its Cloud Security AI Workbench, marking a significant advancement in applying generative AI to enterprise cybersecurity challenges. This comprehensive platform leverages Sec-PaLM, a specialized version of Google's Pathways Language Model that has been specifically fine-tuned for security applications. The introduction of Sec-PaLM represents part of a broader industry shift toward AI-augmented security operations, reflecting how large language models are being adapted to address domain-specific challenges. As organizations face increasingly sophisticated threats and overwhelming data volumes, these specialized AI tools aim to enhance human analysts' capabilities rather than replace them. Understanding Google's approach provides valuable insights into how AI is reshaping enterprise security strategies and what capabilities may soon become standard across the industry.

Allon Oded, VP Product
May 1, 2023

Fake Images of Trump’s 'Arrest' Go Viral, Highlighting Deepfake Danger

In March 2023, a series of hyper-realistic images depicting former U.S. President Donald Trump being arrested by police officers circulated widely across social media platforms. Despite their convincing appearance, these images were entirely fabricated using advanced AI text-to-image generation tools. The timing of these deepfakes was particularly significant, coinciding with a New York grand jury's deliberation on evidence in a criminal case involving Trump. While the former president had publicly predicted his imminent arrest, no such event had occurred at the time the images went viral. This incident represents a troubling milestone in the evolution of synthetic media – demonstrating how easily AI-generated content can infiltrate public discourse, blur the lines between fact and fiction, and potentially influence political perceptions on a mass scale.

Niv Amitay, AI Generation Team Lead
March 21, 2023

Europol Report: How AI Language Models Are Transforming Cybercrime Tactics

The rapid advancement of artificial intelligence has ushered in a new era of technological capabilities. However, as these tools become more sophisticated and accessible, they present significant security challenges that demand our attention. Europol, the European Union's law enforcement agency, recently issued a warning about the potential misuse of large language models (LLMs) like ChatGPT for criminal activities. Their research reveals how these AI tools could lower barriers to entry for cybercrime, creating new vulnerabilities that organizations and security professionals must address proactively. This development marks an inflection point in the ongoing evolution of cyber threats and necessitates a thoughtful examination of both the risks and potential countermeasures.

Doron Ish Shalom, Head of Strategic Partnerships
March 29, 2023

How Cisco Secures Webex Meetings from Fraud

At Clarity we look at the evolution of AI Phishing and how to defend against emerging threats. While we deploy our own solutions to detect and mitigate phishing attacks, we are also encouraged by the efforts of other organizations. Today we look at how Cisco is taking steps to secure their video conferencing platform, Webex.

Allon Oded, VP Product
January 16, 2023

AI threats are evolving.
Your protection should too.