[AI Hanip News] Big Tech Thrived on AI, Will It Fail Because of AI? ... New Risks Appearing in Financial Reports

From Reputation Risks to AI Regulation
MS, Google, Nvidia, and Others Warn of AI Dangers

Financial reports from major U.S. big tech companies such as Microsoft (MS) and Google have included warnings about risks related to artificial intelligence (AI). These companies have benefited greatly from the generative AI boom. However, by mentioning potential risks associated with AI, they appear to be trying to curb excessive expectations.

Key big tech companies including MS, Google, Meta, Nvidia, and Oracle incorporated AI-related risks in their reports submitted this year to the U.S. Securities and Exchange Commission (SEC). While they had previously pointed out climate change, war, and changes in the global financial environment as potential risk factors, they have now included AI risks as well.

Meta and Google, in particular, warned that AI could generate harmful or illegal content. This could lead to the spread of misinformation, bias, or increased cybersecurity threats. Meta specifically noted that ahead of this year’s U.S. presidential election, AI could be used to create false information during the election period. They analyzed that this could lead to backlash against AI models.

Google acknowledged the limitations that the company cannot identify or resolve all potential problems related to AI. They also stated that this could potentially lead to regulatory scrutiny or damage to the company’s reputation. Therefore, they emphasized the need for significant investment. MS echoed these views, pointing out that AI could cause legal issues, citing the European Union’s AI Act and the U.S. AI executive orders as examples.

Some of the concerns raised by companies have already materialized. For example, MS and OpenAI have been embroiled in lawsuits with multiple media outlets, including the Center for Investigative Reporting (CIR), a U.S. nonprofit investigative journalism organization, and The New York Times, over allegations that their AI models infringed copyrights. Some companies have even been blocked from data collection. Meta was ordered to stop using user data for AI training in Brazil following the European Union’s directive. Failure to comply could result in a daily fine of 50,000 Brazilian reais (approximately 12.2 million KRW).

Errors caused by AI have also led to declines in corporate value. Google is a representative case. Its multimodal AI model Gemini caused errors such as depicting U.S. founding fathers and Einstein as people of color and portraying German Nazi soldiers as Asians, which led to Alphabet, Google’s parent company, losing nearly 10 billion KRW in market capitalization at one point.

Nvidia, which has enjoyed a heyday thanks to generative AI, is no exception. In its financial report, Nvidia mentioned that increasing AI regulations could raise costs or disrupt company operations. In fact, Nvidia was affected when the Biden administration banned AI chip exports to China last year.

Oracle and Adobe raised concerns about intensified competition. Adobe pointed out that the spread of AI could shake market demand for existing software (SW) and personnel, while Oracle viewed the risk that its AI products could be outcompeted by rivals. Meta also recognized competitive pressure among AI technology companies as a challenge, even stating that they "may not succeed" in this competition. Additionally, companies like Dell and Uber have newly added AI as a risk factor.