Foreign AI election interference claims under Online Safety Act probe, ministry tells Parliament

Malaysian authorities have pledged to utilize existing laws to counter alleged foreign-linked artificial intelligence influence campaigns targeting Malaysian voters, while simultaneously drafting a dedicated AI Governance Bill for regulatory oversight. This response was provided in a parliamentary reply to an inquiry regarding steps taken by the government to investigate foreign election meddling facilitated through generative AI platforms, such as Claude.

The Communications and Multimedia Commission (MCMC) is treating these allegations seriously and has requested detailed technical clarification and supporting evidence from Anthropic, the US-based creator of Claude, to substantiate any regulatory assessment. Potential enforcement actions will be pursued under the Communications and Multimedia Act 1998 and the Online Safety Act 2025, alongside coordination with federal security bodies, criminal investigators, and social media platforms.

The issue arose following a threat intelligence report by Anthropic in September, revealing that a commercial entity based in Istanbul, BBS Bilisim Teknolojileri, had employed Claude to create core elements of a covert political influence campaign. This operation aimed to construct synthetic persona networks, automated bot accounts, and propaganda websites, and allegedly gathered demographic census records and voter data to profile electorates across all 222 federal parliamentary constituencies.

However, the Turkish company denied any involvement on September 11, and Anthropic confirmed it had shut down the implicated accounts. Despite these findings, it remains unclear whether Anthropic has responded to MCMC's queries, if the police have initiated a formal criminal investigation, or how Malaysian domestic statutes can hold offshore technology operators accountable across international borders.