AI startups step up federal lobbying efforts

AI lobbying at the US federal level is intensifying amid an ongoing productive AI boom and an election year, which could influence future AI regulation.

New data from OpenSecrets, a nonprofit group that tracks and publishes metrics on campaign financing and lobbying, shows that the number of groups lobbying the federal government on AI-related issues has grown from 459 in 2023 to 556 in the first half of 2024 (from January to July). At the same time, top AI startups have ramped up their lobbying initiatives, OpenSecrets data shows.

ChatGPT maker OpenAI has dramatically increased its lobbying expenditures, spending $800,000 in the first six months of 2024, compared with $260,000 in 2023. Meanwhile, the company has grown its team of outside lobbyists from three consultants last year to about 15 in the first half of 2024.

In March, shortly before OpenAI added ex-NSA Director Paul Nakasone to its board of directors, the startup hired former Republican senator Norm Coleman to advocate on research and development issues. Other major law firms, including Akin Gump Strauss Hauer & Feld and DLA Piper, have registered lobbyists for OpenAI, according to OpenSecrets.

OpenAI has also bolstered its internal policy team, hiring Chan Park, a former senior director of congressional affairs at Microsoft, last November to lead its US and Canada partnerships. According to the Financial Times, the number of employees in OpenAI’s global affairs division has more than quadrupled since last year to 35 in eight countries, and the company plans to expand the division to 50 by the end of this year.

On the other hand, OpenAI’s rival Anthropic is set to spend half a million dollars on lobbying in the next few months. So far in 2024, Anthropic has invested $250,000 in its team of five lobbyists – the same amount it spent on three lobbyists in 2023 ($280,000).

Anthropic hired two outside lobbying firms last January, retaining former AWS lobbyist Stoney Burke of Acquia Group and Jed Bhuta of Tower 19. The company also has an in-house lobbyist, who was brought on in early March.

Even small AI companies are spending thousands of dollars on lobbying efforts.

In the first half of this year, Cohere, which invested $70,000 million in lobbying policymakers last year, increased its spending to $120,000, according to OpenSecrets data. Cohere primarily builds custom generative AI models for enterprise customers, a narrower business than OpenAI or Anthropic.

It is no coincidence that lobbyists are getting more support from AI vendors.

For one thing, it’s an election year — and the major presidential candidates have made clear their differing stances on AI regulation.

Former Vice President Kamala Harris of the Democratic Party has indicated that she agrees with President Joe Biden that AI should be subject to some form of federal oversight. On the other hand, former President and Republican candidate Donald Trump has pushed to overhaul the White House’s AI policies and push for general regulation.

The US Department of Commerce released a report this week that may reflect the direction of the Harris administration. The report from the Commerce Department’s National Telecommunications and Information Administration advocates the release of new generative AI models, particularly “open-weight” models like Meta’s Llama 3.1, but recommends that the government develop “new capabilities” to monitor such models for risks.

Congress has yet to pass comprehensive legislation on AI — or even propose legislation as comprehensive as regulations like the EU’s AI Act. The void in federal rulemaking has led to state and local governments rushing to fill the void; nearly 400 state-level AI laws have been proposed this year, according to lobbying group TechNet.

OpenAI, for example, has become more vocal about which AI laws and regulations it prefers, this week supporting Senate bills that would establish a federal rule-making body for AI, provide federal scholarships for AI research and development, and establish AI educational resources in colleges and K-12 settings. (OpenAI has many education clients.)

As the nation awaits the results of the November election, OpenAI and other AI vendors are facing potential antitrust cases from US regulators, including the Department of Justice and the Federal Trade Commission (FTC). CNBC reports that the FTC is seeking more information about Amazon’s recently announced partnership with AI startup Adept and that both the Department of Justice and the FTC are investigating Microsoft’s acquisition of Inflection’s employees. Microsoft gave up an observer seat on OpenAI’s board in July, a move likely aimed at easing the concerns of US antitrust regulators, as Microsoft is a major investor in the company.