Experts Warn AI Could Make Cyber Crime Worse - 2nm.com.au

As artificial intelligence (AI) technology advances rapidly, cybersecurity experts are warning that AI could be exploited to exacerbate cybercrime. Australia's top cybersecurity agencies, in collabora... Read original →

AI-powered cyber threats are triggering a paradigm shift in semiconductor security architecture. Technically, rule-based firewalls and static encryption chips can no longer counter AI-generated dynamic attack vectors, forcing SoC designers to embed reconfigurable AI accelerators and hardware-grade Trusted Execution Environments (TEEs). Regulatory pressure from the Five Eyes alert signals imminent global frameworks against AI misuse, compelling enterprises to adopt zero-trust architectures—raising chip procurement and validation costs by 15–20%. Strategically, NVIDIA and AMD are bundling AI-driven security inference modules with GPUs, while TSMC and Taiwan, China’s supply chain face intensified IP-protection audits. Over the next 12–24 months, adversarial AI will catalyze a dedicated security co-processor market (CAGR >35%) and accelerate RISC-V’s secure extension instruction set as an industry standard.