Can You Spot a Deepfake? Malaysians Share Their Thoughts on AI-Generated Content
Nearly nine in 10 Malaysians have seen content they suspect is AI-generated, according to a live poll conducted by MIMOS Berhad at the National Dialogue on Trusted Intelligence.
Source: Malaysian Institute of Microelectronic Systems · August 12, 2026 at 7:41 PM · AI-assisted report
Single-source
KUALA LUMPUR, 13 AUGUST 2026 —
Nearly nine in 10 Malaysians have seen content they suspect is AI-generated, according to a live poll conducted by MIMOS Berhad at the National Dialogue on Trusted Intelligence.
Market Impact
The survey of 105 government, industry, academia and technology leaders held on 8 July 2026 found 85% of participants reported encountering suspected deepfakes. That compares with 64% who said they regularly verify sources before believing or sharing posts.
Participants described deepfakes as a growing national security threat, with 75% labelling them a significant risk to Malaysia’s digital infrastructure. The responses reflect a broader shift in public perception beyond simple fraud, said MIMOS, a government-backed applied research agency under the Ministry of Science, Technology and Innovation.
“Deepfakes are no longer a futuristic concept,” a MIMOS spokesperson said. “They are today’s reality, and the technology is advancing far faster than our ability to detect it.”
The poll captured real-time reactions on sentiment around AI-generated content. Words volunteered by participants—fraud, deception, crime, risk and misinformation—showed a predominantly negative view, even as the same tools power creative uses such as animated humour and photo-to-video conversions.
MIMOS hosted the dialogue to gather industry perspectives on digital trust. The event was held amid rising reports of AI-enabled financial scams and identity theft across Malaysia’s social platforms. Participants agreed that the gravest risk is erosion of trust: when manipulated content becomes indistinguishable from genuine, people begin to doubt even verified information.
The survey asked who is most vulnerable to deepfake-based crimes. Elderly Malaysians were cited most often, followed by small business owners and low-digital-literacy groups. Combating the threat requires layered defences, participants said: faster detection tools, tighter regulations, and sustained public awareness campaigns.
Bank Negara Malaysia has flagged deepfakes as a risk to financial stability, citing rising cases of voice-cloned loan scams reported to commercial banks. Local chip designers, data-centre operators and cybersecurity firms already supply regional AI infrastructure, yet domestic detection capabilities remain fragmented.
Participants highlighted three mechanisms they trust most: verification from official sources, fact-checking labels on platforms, and government confirmation of major announcements. “Trust should be built through verification, not assumptions,” said an academic from Universiti Teknologi Malaysia.
Industry executives say Malaysia’s regulatory sandbox for AI could accelerate adoption of detection standards. “We need to move from reactive takedowns to proactive watermarking,” said the chief executive of a Kuala Lumpur-based cybersecurity firm. “Malaysia can set the template for Southeast Asia if we align early.”
MIMOS said it will publish the full survey data next month alongside a draft framework for public consultation. Until then, participants urged every Malaysian to pause before sharing: question the content, verify the source, and share responsibly.
Related: Intel