Singapore (VNA) – Singapore’s Cyber Security Agency (CSA) has urged businesses to strengthen cybersecurity measures amid growing concerns that advanced artificial intelligence (AI) models could increase cyberattack risks.
The warning came on April 15, shortly after US company Anthropic unveiled a preview of its next-generation AI model, Mythos, which has drawn attention for its powerful capabilities. The UK’s AI Security Institute found that the model could potentially assist in carrying out more sophisticated cyberattacks than widely used AI tools such as ChatGPT or Gemini.
According to CSA, advanced AI models – among the most cutting-edge systems today – can significantly shorten the time needed to identify system vulnerabilities and develop exploitation tools, reducing the process from months to just hours.
In response, CSA has advised businesses to immediately adopt risk-mitigation measures, including patching vulnerabilities in internet-connected systems, implementing multi-factor authentication, and reviewing access permissions to remove unnecessary privileges.
The agency also stressed that advanced AI represents a major step forward for cybersecurity, as it can analyse large volumes of source code and detect weaknesses much faster than manual methods. However, the same capabilities could be exploited by malicious actors to accelerate vulnerability discovery and develop cyberattack tools.
CSA said there is currently no evidence that such capabilities are being misused but issued the advisory to help organisations proactively prepare and strengthen their cyber defences.
In the longer term, businesses are encouraged to develop comprehensive risk-mitigation strategies, including monitoring and reducing potential attack paths and deploying multiple layers of security.
Meanwhile, Anthropic said that during the development of Mythos, the system detected thousands of critical vulnerabilities, including weaknesses in major operating systems and web browsers.
The development has raised global concerns, prompting financial regulators in the UK and the US Treasury to discuss potential risks posed by emerging technologies./.
