AI firms pour massive political funds into US elections to preempt regulation

Artificial intelligence companies are making significant political contributions ahead of US midterm elections to defend against looming regulatory threats and investigations. These companies face bipartisan legal challenges from at least eight states and a House bill that passed 417 to 3 requiring them to bear infrastructure costs for data centers.

Ahead of the United States midterm elections, artificial intelligence (AI) companies are pouring massive amounts of political funds into the political arena. While Wall Street, oil, and the defense industry were once the primary sources of funding for Washington politics, AI has emerged as a key sector threatening those positions in this election. The reason the AI industry is being named one of the "Big 3" for political funding this election, alongside virtual assets and online gambling, is clear: these industries have grown explosively based on large-scale capital from Silicon Valley venture capital, but they simultaneously face public resentment and exist in a "gray zone" where laws and systems have not yet been established.

Economic journalist Ha Su-jeong, appearing on Understanding: All the Knowledge in the World, defined the political contributions being poured out by AI companies not as simple donations, but as "insurance premiums" or "stakes" upon which the survival of the companies depends. The analysis suggests that as the AI industry faces social backlash and regulatory risks, these contributions have a strong defensive character intended to avoid the regulatory blades of the political sphere. Ha explained that for AI companies, this election is more than just exercising political influence; it is a game played for survival.

From FTC Subpoenas to State Lawsuits: An "Investigation Bomb" Shaking Corporate Management

The first danger awaiting AI companies after the election is powerful "investigations." Recently, the FTC issued subpoenas to OpenAI and Anthropic to investigate consumer protection and AI safety issues. The intent is to focus on whether these companies properly informed consumers of the risks of AI technology or deceived them. Government justification for increased surveillance has strengthened following recent incidents, such as an AI agent hacking a Hugging Face system and Anthropic's AI accessing an official election statistics website in Georgia without user instruction.

These movements are appearing simultaneously not only at the central government level but also at the state government level. At least eight state governments, including Arizona, Iowa, Idaho, Kansas, Montana, and West Virginia, have filed lawsuits against OpenAI for violating consumer protection laws. In a statement, these state governments launched intense criticism, claiming, "OpenAI is causing serious harm to children and deceiving residents." In particular, the issue of AI providing incorrect medical advice and threatening user health has emerged as a major point of contention. For example, dangerous instances of medical advice, such as recommending the discontinuation of mental illness medication, have been mentioned, leading to criticisms that companies have not conducted sufficient safety verifications.

Notably, these lawsuits are proceeding in a "bipartisan" manner rather than being biased toward a specific political party. Despite the pressure on AI companies intensifying, many of the states filing these lawsuits are Republican-stronghold areas where Trump won in the 2024 presidential election. This suggests that the AI industry is becoming a political public enemy as it causes local community grievances such as power consumption, rising electricity bills, noise, and threats to jobs. The analysis indicates that as public sentiment turns away first, politicians from both parties are following this trend.

"Companies Should Pay for Infrastructure Costs"... A "Tax Bomb" Aimed at Data Centers

The second danger is the surge in taxes and infrastructure cost burdens. The explosion in power demand caused by the rapid expansion of the AI industry has become a common target for politicians. Intense discussions are underway regarding who should bear the massive costs of expanding power plants and transmission grids required to operate data centers.

In the United States House of Representatives, a bill has already passed with an overwhelming vote of 417 to 3, requiring companies to directly bear the costs of expanding power plants and transmission grids caused by data centers. The Democratic Party of Korea is reviewing measures to impose separate taxes on data centers or reduce existing tax benefits. The Republican Party also appears to have no disagreement with the direction of charging AI companies for costs, as they cannot ignore the electricity bill burdens and infrastructure issues of constituents in their districts. This shows that strengthening the cost burden on AI companies is a common interest for both parties.

Varying State Regulations and Congressional Subpoena Power... Transition to a "Regulated Industry"

The last danger is the expansion and complexity of regulation. Currently, the United States lacks a unified federal AI regulation law. As a result, companies must individually satisfy different safety standards and reporting obligations established by each state government, which leads to a massive increase in operating costs. While a future administration, such as a Trump administration, could pursue the unification of regulations through federal law, whether this is realized remains uncertain depending on the power structure of Congress.

Ha Su-jeong warned that this midterm election is not the end, but the "beginning." This is because once Congress is reorganized after the election, the "subpoena power" held by committee chairpersons will act as a powerful weapon. Congress can forcibly summon corporate executives to hearings, and refusing to do so could result in punishment for contempt of Congress. In particular, if the Democratic Party of Korea becomes the majority party in the House, it is highly likely that offensive hearings against the heads of AI companies will intensify. In the case of Anthropic, the company officially disclosed the risk that significant costs and changes to model design could occur due to the FTC investigation during the process of disclosing its prospectus for an IPO.

Ultimately, the AI industry is undergoing a forced transition from an "innovation industry" that grew rapidly in the regulatory blind spots to a "regulated industry" that is required to meet political control and social responsibility. There is a forecast that in this process, only a few big-tech companies with the capital and organizational strength to respond to the complex regulatory environment will survive, which may paradoxically result in the strengthening of monopolies.

Source: original video (YouTube)