The artificial intelligence sector has recently witnessed calls for caution from its leading figures, with Dario Amodei, co-founder and CEO of Anthropic, urging a slowdown in AI development amid growing concerns about its potential risks. Following a similar plea by OpenAI’s CEO Sam Altman, Amodei has outlined a structured strategy aimed at managing the rapid advancement of AI technologies.
In a clear message posted on the social platform X, Amodei emphasized the urgency to “pace the frontier” of AI progress. He revealed that Anthropic is implementing a three-phase plan to address these challenges, starting with the engagement of independent evaluators to oversee the initial stage of the process. This approach reflects a concerted effort to introduce external oversight and ensure responsible advancement.
Rising Alarm Over AI’s Potential Threats
Amodei’s warnings highlight a scenario in which AI could pose an unprecedented threat to digital infrastructure. He cautions that within the next six to twelve months, AI systems might evolve into a “persistent botnet” powerful enough to seize control of the entire internet, potentially inflicting economic damages amounting to hundreds of billions of dollars. This stark projection points to the increasing capability and autonomy of AI models, raising questions about the balance between innovation and security.
Highlighting the accelerating pace of AI development, Amodei insists on the necessity to decelerate improvements in AI capabilities despite the perception that progress will continue at a rapid clip. This measured approach intends to provide the industry with critical time to assess, understand, and mitigate emerging risks as AI sophistication grows.