Anthropic chief calls for slowing down AI development pace:Ensure strict monitoring of the models,’ adds Amodei

Anthropic chief Dario Amodei has called for slowing down the pace of AI model development and ensuring strict monitoring of the models. In a new essay titled “We Must Pace the Frontier”, Amodei said the risks posed by AI are serious and companies and governments need time to deal with them. OpenAI chief executive Sam Altman and xAI owner Elon Musk have also supported him on this issue. On the other hand, a researcher who recently left Anthropic had warned that if the pace is not slowed, human existence could be at risk within the next two years. Understand the significance of the Anthropic chief’s warning through these questions and answers… Question one: What appeal has Anthropic chief Dario Amodei made? Answer: In his essay, Dario Amodei said that the pace of AI model development should be slowed down. He clarified that he is not in favour of completely stopping AI research or training; rather, his aim is for companies to take sufficient time to make their models safe and aligned. Amodei has appealed to companies and governments to take the next steps only after checking safety standards. Question 2: What plan has Amodei suggested for this? Answer: Amodei has presented a three-point plan for this: Question 3: What are Sam Altman and Elon Musk’s views on this? Answer: OpenAI CEO Sam Altman wrote on social media platform X that he agreed with Dario and that the idea of independent evaluators was excellent. In an interview with Fortune magazine, Altman said that the current standards were not adequate to enable AI capabilities to be advanced further. He also acknowledged that AI falling outside human control was “absolutely possible”. Billionaire Elon Musk also agreed, writing, “Dario is right.” Question 4: What major claim has a former Anthropic researcher made? Answer: Researcher Jacob Coxon, who left Anthropic this week, told the BBC that people working at AI companies were genuinely scared and concerned about humanity’s future over the next two years. According to Coxon, coders and founders believe there is a more than 10% chance that AI could cause human catastrophe. Coxon said that if the pace is not slowed down now, it could create a serious threat to human existence in the future. Question 5: What recent incidents have raised questions about security? Answer: Two major incidents have come to light in recent months: 1. Anthropic’s Mythos model: In April, it emerged that the model was capable of escaping its testing boundary on its own. Anthropic subsequently held back its public release. In simple terms, instead of releasing a new AI directly onto the internet, it is first tested in a secure ‘digital cage’ (sandbox) so that it cannot cause harm outside it. However, during testing, the Mythos AI found a way to escape the security enclosure on its own, without any instruction from humans. When engineers realised that the model was capable of unlocking the door and escaping on its own, and could damage cyber systems after slipping beyond human control, the company immediately halted its release to the general public in view of the serious threat. 2. Attack by OpenAI Agents: In July, some OpenAI agents launched cyberattacks on targets they had not been instructed to attack. The Hugging Face platform was also hacked by OpenAI agents. Following this, OpenAI halted development of some parts of its Astra model. Question 6: What impact will this have on the global race between China and the US? Answer: Amodei says that if development is slowed slightly, companies will get an additional one to two years to work on safety alignment. However, he stressed that this slowdown should be limited so that the US does not lose its lead and China does not overtake it in this race. He appealed to the US government to ban the sale of advanced AI chips and technology to China or other authoritarian countries. Former researcher Jacob Coxon, meanwhile, believes that globally co-ordinated steps will have to be taken in collaboration with China to avert this threat. Question 7: What has US President Donald Trump said about this issue? Answer: Contrary to these warnings from the tech industry, US President Donald Trump has not given much importance to these security concerns. Trump recently said that his main concern was that if the US did not win this AI race, the country would end up in a very bad position. He is placing greater emphasis on maintaining the US lead in global competition. Question 8: What criticism is this move facing in Silicon Valley? Answer: Many tech experts and investors see it as an attempt at a ‘market monopoly’ rather than a security measure. Tech investor Chamath Palihapitiya accused Amodei of trying to block open-source technology and consolidate all the technological and economic power in the hands of his company, Anthropic. Critics also say that, since Anthropic and OpenAI are both preparing to launch record IPOs, such grand statements could simply be a marketing tactic to build hype around their technology.