MCMC seeking clarification from Anthropic over election manipulation claim

KUALA LUMPUR: Malaysian authorities are requesting clarification from Anthropic regarding allegations that an artificial intelligence (AI) system called Claude was utilized to manipulate Malaysian elections. Communications Minister Fahmi Fadzil stated that the Malaysian Communications and Multimedia Commission (MCMC) is collecting information to decide whether additional steps are required to address the situation.

This inquiry aims to ensure that any subsequent actions are grounded in sufficient data and evidence, according to Fahmi's written parliamentary response. Anthropic contends that an overseas entity employed Claude to create elements of a political influence campaign, encompassing voter-targeting mechanisms, networks of fictitious social media personas, and fictitious news outlets.

The company identified these components as being linked to BBS Bilisim Teknolojileri, a technology firm situated in Istanbul, Turkey. Following its identification, Anthropic reported that it had deactivated the implicated Claude accounts and implemented countermeasures. The Malaysian government has pledged to enforce relevant regulations, such as the Communications and Multimedia Act, the Online Safety Act, and other pertinent statutes within MCMC's purview.

The Digital Ministry is concurrently drafting an Artificial Intelligence Governance Bill to bolster Malaysia's AI regulatory landscape, focusing on accountability, risk management, and the obligations of entities creating and deploying AI systems. In September, Anthropic disclosed the takedown of an alleged foreign-backed election manipulation effort that targeted Malaysia with approximately 1,000 fictitious social media profiles, a fabricated news website, and false intelligence reports.

MCMC has asserted that it is not implicated in the described activities and is currently investigating the allegations to ascertain the veracity of the claims.