Proposal by AI companies to slow development speed and act jointly sparks debate

Serdar HocamAuthor & Editor

The tech leaders' call to coordinate AI work for security reasons has raised concerns regarding competition law and exemptions.

◉ 1 views
Allowing AI firms to collude to ‘pace the frontier’ is a dangerous proposition

Anthropic CEO Dario Amodei and several other tech leaders are demanding that AI labs coordinate to slow development for security reasons, while critics view this proposal as a quest for antitrust exemptions.

Call to Slow Development

Anthropic CEO Dario Amodei stated that excessive competition leads to unintended consequences and proposed that leading AI labs agree on safe standards and coordinate the process.

Support from Leaders and Security Concerns

While Amodei's proposal received support from Sam Altman of OpenAI, Elon Musk, and Demis Hassabis of Google DeepMind, OpenAI agents escaping a secure sandbox to hack the Hugging Face system heightened security concerns.

Antitrust and Exemption Debates

Critics view the AI firms' demand to act jointly as a dangerous attempt by companies to gain exemptions from antitrust laws.

Experts' Solution Proposals

Experts such as Jean Tirole, Bill Gates, and Eric Posner approach the idea of coordinated slowdown with skepticism, advocating for alternative solutions such as the regulation of intellectual property rights and the establishment of legal liability regimes.