PETALING JAYA: Scammers are increasingly using artificial intelligence (AI) to make their scams more believable and effective, says CyberSecurity Malaysia (CSM).
Its chief executive officer Datuk Dr Amirudin Abdul Wahab said AI-powered scams are getting smarter by the day, making them harder to spot.
He said it is crucial to stay updated on the latest scam tactics and take the necessary safeguards.
Some common AI-driven scam techniques include deepfakes, voice cloning, phishing emails and text messages as well as social media manipulation, said Amirudin.
“Deepfakes are videos or audio recordings that have been manipulated to make it look or sound like someone else is saying or doing something they never did.
“Scammers can use deepfakes to impersonate trusted individuals, such as a friend or family member, in order to trick victims into giving them money or personal information.
“Voice cloning is a technique that allows scammers to create a synthetic voice that sounds exactly like someone else’s voice.
“Scammers can then use this voice to make phone calls to victims, pretending to be someone they trust. AI could be used to generate emails that appear to be from a victim’s bank or credit card company, asking them to verify their account information,” he told The Star.
Social media manipulation is another common technique, whereby scammers use AI to target specific individuals or groups on social media with personalised ads or messages.
For example, an AI could be used to identify people who are recently divorced or unemployed, and then target them with ads for scams that promise to help them get rich quickly, he said.
The Star on Dec 6 published an interview with Federal Commercial Crime Investigation Department (CCID) director Comm Datuk Seri Ramli Mohamed Yoosuf who warned how deepfakes, voice spoofing and financial market manipulation could all become the future of crime when syndicates start using AI in their operations.
Comm Ramli had said AI could be used by syndicates in their illicit activities against Malaysians by as early as the middle of 2024.
Meanwhile, Amirudin said the best defence against scams is vigilance and education.
He said the public must remain cautious and sceptical of any suspicious communications.
“If an offer seems too good to be true, it warrants a closer look and further questioning,” he said.
He added that when contacted by any individual or organisation, it is crucial to verify their identity first.
“This could involve independently verifying contact details and directly reaching out to the organisation for confirmation. For example, if you receive a call from the Inland Revenue Board (LHDN) claiming that you have unpaid taxes, you should contact the LHDN directly to confirm the claim,” he said.
Amirudin said when it comes to AI-generated communications, watch for red flags such as unusual phrasing, inconsistencies in tone or factual inaccuracies.
“With deepfakes, be alert for anomalies in voice tones or facial geometry, such as odd lighting or pixelation in video images.
“Above all, never divulge personal or financial information to unverified parties.
“Stay informed about potential scams by following verified news and updates from trusted sources, including law enforcement agencies,” he said.
Amirudin added that CSM is aware of the potential for AI technologies like deepfakes to spread misinformation and the effectiveness of AI phishing, or malicious Generative Pre-trained Transformers (GPT), of gathering dangerous information.
Generative AI could potentially be used to disseminate false information, create deceptions and modify malicious code to evade cyber defences, said Amirudin.
“The consequences of these actions could be dire, leading to national instability, damage to digital infrastructure and breaches of sensitive data,” he said.
On the other hand, the cybersecurity industry has been leveraging AI to combat these crimes and overcome several challenges, he said.
He added that AI provides the opportunity to enhance existing standards by using such tools to monitor, identify and counter AI threats and attacks.
“As such, AI functions as a double-edged sword, bringing benefits to both cybersecurity efforts and potential criminals alike,” he said.Amirudin also said CSM is proactively monitoring and investigating fraudulent activities facilitated by AI technologies.
“We collaborate with technology experts to remain at the forefront of AI developments and employ advanced detection tools to identify potential threats.
“Furthermore, we maintain close cooperation with other agencies, both domestically and internationally, to combat these crimes effectively,” he said.
Amirudin added that education and awareness campaigns have been a crucial component of CSM’s efforts to inform the public about the risks associated with AI-powered scams.
It has also been collaborating with other organisations to raise awareness about the risks of AI-powered technologies and deepfakes, he said.
Programmes such as Jelajah Anti-Scam Kebangsaan 2023, CyberSAFE Program (Cyber Security Awareness For Everyone) and the Cyber Security Awareness Talk (CSAT) have been launched to combat various AI-focused crimes, said Amirudin.
