S’pore steps up fight against deepfake scams: Gov’t and industry collaboration key

SINGAPORE: Singapore authorities are stepping up efforts to combat the nefarious use of deepfake technology in scam and fraud cases, with a slew of regulations and initiatives aimed at safeguarding citizens against online harms.

Mrs Josephine Teo, Minister for Communications and Information, explained this during a parliamentary debate on Monday (5 February).

Her remarks came in response to inquiries from Mr Christopher de Souza regarding the strategies being implemented to address the prevalence of deepfake software in scam and fraud incidents.

Additionally, Mrs Teo elaborated on measures aimed at augmenting public technological literacy to combat cybercrime and deepfake content.

Deepfakes refer to media that has been altered by artificial intelligence (AI), using a technique called facial re-enactment to manipulate the performance of a subject in an existing video.

S’pore steps up efforts against deepfake scams: Government and industry collaboration key

Mrs Teo began by stating that AI has the potential to be misused for malicious intents, such as perpetrating scams and fraud.

As such, various agencies such as the Ministry of Communications and Information (MCI), the Ministry of Home Affairs (MHA), the Singapore Police Force (SPF), the Infocomm Media Development Authority (IMDA), and the Cybersecurity Agency of Singapore (CSA) are working closely to ensure that Singaporeans can go online safely and safeguard themselves against such online harms and threats.

One of the key strategies Mrs Teo mentioned involves holding social media platforms accountable for swiftly removing scam content and preemptively detecting and blocking deepfake-enabled scams.

Under the Online Criminal Harms Act (OCHA), passed in July 2023, the government has the authority to issue directives to online platforms to prevent potential scam-related content from reaching Singapore users.

Mrs Teo said that, “under OCHA, designated online service providers may also be required to implement measures (if not already taken) to proactively disrupt online scams, including those facilitated by deepfakes.”

“The Government is working with industry partners to strengthen our capabilities to deal with these threats,” she added.

Several of these efforts were also discussed during an earlier Parliamentary Motion on Building an Inclusive and Safe Digital Society.

One example is the forthcoming Centre for Advanced Technologies in Online Safety, which seeks to promote collaboration and knowledge sharing within the industry to detect deepfakes effectively.

Other than that, the SPF, in collaboration with the Home Team Science and Technology Agency, is also working on developing and enhancing technologies to detect AI-generated audio and videos.

This effort aims to strengthen the nation’s defenses against the harmful exploitation of deepfake technology.

Government enhances public education for a safe digital society

In addition to regulatory measures, the government is intensifying public education efforts to promote digital media literacy, cybersecurity awareness, and scam prevention.

These measures is to “complement the Government’s effort to build a safe and inclusive digital society,” Mrs Teo said.

Signature campaigns such as the National Library Board’s S.U.R.E. (Source. Understand. Research. Evaluate) campaign, the CSA’s “Unseen Enemy” cybersecurity campaign, and the SPF/National Crime Prevention Council’s “I can ACT against scams” initiative are aimed at empowering citizens to navigate the digital landscape safely.

Additionally, The  Scam Public Education Office was also set up in 2023 to drive anti-scam public education efforts and expand outreach.

Mrs Teo said, “the Government will closely monitor and continue to adjust our strategies and tools to keep pace with the rapidly evolving technological landscape.”

Surge in deepfake cases

In 2023, Singapore witnessed a fivefold surge in deepfake cases, marking a significant increase within that year alone.

The Sumsub Identity Fraud Report 2023, unveiled in November, revealed a tenfold rise in globally detected deepfakes across various industries from 2022 to 2023.

Addressing the challenge posed by deepfakes, Professor Mohan Kankanhalli, Deputy Executive Chairman of AI Singapore, described it as “extremely challenging” in an interview with Channel NewsAsia (CNA).

Professor Kankanhalli noted that the detection of early-generation deepfakes was relatively simple, given noticeable imperfections like unblinking eyes. However, he highlighted that scammers have since identified and rectified these flaws, making the software more sophisticated.

As a result, generative AI technology continues to advance rapidly over time, with Prof. Kankanhalli emphasizing the ongoing enhancement of these capabilities.

He stressed the importance of regulators comprehending the technology, pointing to Singapore’s IMDA as an actively engaged entity in the AI landscape, well-informed about advancements in the field.

Mrs Teo during a parliament debate last month (10 January) had also suggested that Singapore is among top countries for scam amounts due to the high reporting levels by victims.

Despite Singapore’s leading position in this regard, she highlighted a widespread reluctance among scam victims worldwide to report these incidents.

Mrs Teo observed that many victims refrain from reporting scams due to a lack of trust in authorities’ capability to handle these matters effectively, leading to underreporting across different regions.

Nonetheless, she also highlighted that Singapore differs markedly from this pattern, exhibiting notably high levels of reporting.