Cybersecurity and Scam Prevention

Deepfakes and the Crisis of Credibility

9 January 2026

How can Singapore remain resilient against digital deception? Discover how the nation addresses the rising risks of deepfakes through strong safeguards and public awareness, protecting trust in the digital age of synthetic media.

Image featuring abstract icons of human faces and facial recognition motifs, representing digital deception.

Deepfakes have rapidly evolved into a serious digital threat, eroding the line between fact and fiction. By fabricating convincing audio, video, and images, this technology poses far-reaching risks for society, making public awareness critical. In Singapore, staying ahead of this fast‑moving “digital arms race” is a national priority, as synthetic media makes it harder to distinguish fact from fiction.

What Makes Deepfakes a Growing Threat?

Deepfakes are synthetic media generated by artificial intelligence (AI) that manipulate or create realistic images, audio, or video. In simpler terms, a deepfake is a highly convincing fake, often appearing genuine, created using AI to make someone appear to say or do something they never did. Deepfake technology itself can have legitimate uses in entertainment or education. However, the danger of it lies in their ability to exploit our natural tendency to trust what we see and hear. When deployed maliciously, they pose risks ranging from automated scams, such as impersonating executives to authorise fraudulent transfers, to widespread disinformation campaigns that manipulate public opinion. Furthermore, they can cause severe reputational damage by creating non-consensual content used for harassment or blackmail.

The Singapore Context: Scams and Legal Defence

The threat is palpable in Singapore, where deepfake scams are escalating. Recent data indicates a surge in business risk, with 56% of Singaporean businesses reporting incidents of audio deepfake fraud and 52% encountering video deepfakes.

High-profile cases have already underscored the impact on public trust:

Singapore’s legal framework has evolved to address these specific threats. This includes the Elections (Integrity of Online Advertising) Act, which bans the misrepresentation of election candidates via digitally manipulated content, and the Protection from Harassment Act (POHA) for individuals facing distress caused by deepfakes.

Engineering a Proactive Ecosystem: GovTech’s Multi-Layered Defence

Deepfakes are only one facet of a wider challenge: the rise of digital deception. To counter both synthetic media and the industrial scale of modern online fraud, the Government has moved beyond reactive measures to build an ecosystem of proactive, automated defences.

Strengthening Government Integrity with INDEPTH

At the forefront of technical detection is INDEPTH, an integrated deepfake detection hub designed specifically for government agencies - developed by GovTech in collaboration with the Ministry of Digital Development and Information (MDDI). While deepfakes are increasingly indistinguishable to the human eye, INDEPTH leverages advanced AI algorithms to detect synthetic and manipulated media.

It provides a crucial capability for public officers to analyse audio, videos and images for signs of manipulation—such as pixel-level inconsistencies—thereby safeguarding against misleading content so that Singapore remains resilient against evolving online threats. By automating the extraction of insights from vast volumes of data, INDEPTH allows agencies to identify anomalies and quickly triage suspicious content at machine speed.

In fact, INDEPTH was deployed in support of the Elections (Integrity of Online Advertising) Act for the General Elections 2025.

Hunting Scams with SATIS

Alongside deepfakes, scams remain a major vector of digital deception. The Scam Analytics and Tactical Intervention System (SATIS), a solution developed by GovTech Singapore in collaboration with the Singapore Police Force (SPF) operates as an always-on sentry, integrating threat intelligence to hunt and neutralise malicious infrastructure before it can claim victims.

The Engine: SATIS is powered by a proprietary AI model known as the recursive Machine-Learning Site Evaluator (rMSE). This in-house classifier can analyse website features at machine speed to determine if a site is legitimate or a fabrication.

The Scale and Speed: The system continuously scans over 400,000 websites daily, each within seconds, while achieving a precision rate of over 90%. By automating the detection and assessment of scam websites, the system enables authorities to tackle scam sites at a scale human analysts alone could not match.
The Intervention: When a site is confirmed as malicious—such as a bank phishing page or a fake job portal—SATIS triggers a disruption. This results in the site being blocked at the DNS level by the SPF and displaying a stark red warning from Google, effectively cutting off the scammer's access to potential victims.

SATIS disrupts 20,000 sites a month, most within 10 minutes of detection, effectively preventing more than 10 million scam attempts every month.

Empowering Citizens with ScamShield

While INDEPTH and SATIS equip public officers with advanced detection capabilities, the ScamShield suite arms citizens with real-time protection against the broader spectrum of scams. Established as a joint effort by the Ministry of Home Affairs, the Singapore Police Force, the National Crime Prevention Council, and Open Government Products, it consolidates anti-scam resources into a single, trusted product suite.

Crowdsourced Intelligence: The ScamShield app allows users to report suspicious calls, messages (SMS, WhatsApp, Telegram), links, and emails. These reports help block scam numbers flagged by authorities, protecting users from future attempts.
On-Device AI: To protect user privacy, the app analyses incoming SMSes directly on the device. Suspicious messages are filtered into a “Junk” folder, while messages from recognised businesses or your saved contacts are considered safe.
Holistic Support: Beyond the app, ScamShield offers a 24/7 Helpline (1799) for immediate verification and assistance.

Together, INDEPTH, SATIS, and ScamShield illustrate Singapore’s multi‑layered defence against digital deception — from synthetic media to large‑scale scams. Yet technology alone cannot fully safeguard trust. Deepfakes in particular demand vigilance at the individual level, because even the most advanced systems cannot catch every forgery in real time.

How to Spot a Deepfake: Tips for Digital Hygiene

Stopping deepfakes is a shared responsibility. Identifying sophisticated forgeries often requires critical observation from the user.

When evaluating videos or images, observe these visual inconsistencies:

Feature	What to Look For
Facial irregularities	Unnatural blinking, uneven skin tone, or strange lighting around the face.
Edges & Background	Digital artefacts or a distorted background that doesn't align with the foreground.
Synchronisation	Mismatch between lip movements and the spoken words.

Strengthen Your Personal Defences

Verify the Source: Cross-check "too good to be true" schemes with official channels before acting.
Limit Personal Exposure: Reduce the amount of photos, videos, or voice recordings you upload publicly to limit the data available for cloning.
Enable two‑factor authentication (2FA): Activate two-factor authentication on all sensitive accounts to prevent unauthorised access even if your password is compromised.

Staying Safe in a Synthetic Future

As AI capabilities accelerate, the arms race between fabrication and detection will intensify. By combining personal vigilance with national-level engineering solutions Singapore is building a resilient digital society capable of weathering the deepfake storm.

If you suspect you have encountered a deepfake scam, call the ScamShield Helpline at 1799 for immediate assistance, or visit scamshield.gov.sg for more information.