Nirvana bassist Krist Novoselic led failed shareholder proposal asking Microsoft to study AI safety

Krist Novoselic, co-founder and bass guitarist for the seminal rock band Nirvana, recently led a shareholder proposal presentation urging Microsoft to reevaluate its approach to generative artificial intelligence (AI).

Called Shareholder Proposal 13: Report on AI Misinformation and Disinformation, per a press release, the proposal was submitted by Arjuna Capital “on behalf of Krist Novoselic” and several other shareholder groups.

“Novoselic, co-founder and band member of Nirvana, will present the proposal citing concerns that Microsoft has not fully considered the business and societal risks of generative AI amid its rapid deployment of the technology.”

Proposed artificial intelligence risks

The proposal cited several key shareholder concerns, including the potential for Microsoft-developed or -backed models to participate in the spread of mass disinformation and misinformation.

It also brought up questions surrounding whether Section 230, a law giving internet hosts and website users limited protections against liability for content from third parties, would even apply to content generated by the host’s own generative artificial intelligence systems.

In his presentation, Novoselic asked, “What happens when society relies on information generated through Microsoft’s AI-powered Bing, a platform found to produce inaccurate answers ten percent of the time?” He also brought up the calls for a six-month pause on AI development from experts earlier this year, which Microsoft, alongside the rest of the industry, chose not to heed.

According to Novoselic, Microsoft’s rush to market “seemingly prioritized short term profits over long term success.”

Microsoft’s response

Microsoft’s board responded to the proposal by stating it had “already fulfilled the proposal’s request with existing and upcoming reporting.” However, according to Arjun Capital, the proposal’s purpose was to get “information that goes beyond these reports’ generic commitments to responsible AI.”

Specifically, the shareholders behind the proposal sought to spur Microsoft to comprehensively assess the risk associated with generative AI in the long term.

Citing its current programs and reporting as sufficient, Microsoft’s board made the recommendation that shareholders deny the proposal. The proposal failed to pass a subsequent shareholder vote.

Related: Microsoft faces UK antitrust probe over OpenAI deal structure