Within the technology industry and international political leadership there are different positions. What divides us above all are two issues: the safety of the models and the pace of their evolution. The comparison is now polarized between the enthusiasm of those who intend to accelerate, the alarms of the apocalyptics and the diplomatic caution of the large internet giants. The debate is also heated between those who think that rules are needed to regulate the sector and those who disagree.
Among the big names in technology, Dario Amodei of Anthropic, Sam Altman of OpenAI and Elon Musk of xAI were exposed: despite being engaged in a daily rush for announcements, they asked for caution. On the other side there are Mark Zuckerberg of Meta and Jensen Huang of Nvidia, who believe that legal and engineering responsibility is enough to guarantee safety for those who develop AI, without slowing down progress.
In the background, then, there is politics. The White House, with the Trump administration, has repeatedly reiterated its position, underlining the importance of maintaining the technological primacy of the United States in a sector that is increasingly central to the global scenario. Also because China, Washington's fixed thought, continues to run with massive state investments and aggressive industrial plans aimed at bridging the gap with the West.
On the one hand, therefore, there are those who support advancement at all costs: this is where the figures convinced of the inevitability of progress are concentrated. Nvidia is pushing in this direction, also driven by economic interests given that its semiconductors constitute the material infrastructure of the artificial revolution. "Security is an engineering question, not a legal one," Huang said. “At the end of the day, we are developing software,” he added.
A line that is also reflected in the policy of Donald Trump, who has decided - as mentioned - to eliminate all brakes to preserve Washington's technological advantage over Beijing. US Vice President JD Vance also spoke out on the issue, rejecting the request of the artificial intelligence giants to impose new rules. "If you're making Frankenstein, stop. If you're going to make Frankenstein, don't come to the government and say we need rules," Vance said.
Trump's line, which wants to preserve the advantage over China, also finds Mark Zuckerberg in favour. The CEO of Meta rejected the hypothesis of an agreement between companies to slow down the development of AI, claiming the full autonomy of individual laboratories in managing security and proposing the use of independent auditors instead of collective moratoriums. "There is no need to slow down on AI but to build trust," said the founder of Meta in a post on social media.
On the opposite side, then, we find the most alarmed voices, often originating within the leading companies themselves. Dario Amodei, co-founder of Anthropic, formalized the request for a pause in releases to give developers time to integrate fundamental security requirements without affecting American competitiveness.
In a more complex and nuanced position, however, is Sam Altman of OpenAI: after having warned against the risk of losing control and the dystopian drifts linked to the concentration of technological power, the manager spread messages of reassurance, declaring full confidence in the ability of the sector and his company to develop AI without fear for humanity. According to the Financial Times, however, tensions emerged within OpenAI and Anthropic regarding the positions of the CEOs.
In the middle, then, there are giants like Google and Amazon. The two multinationals - with long experience in US regulatory dynamics - are adopting a more neutral and calculated positioning: according to experts, this is a typical attitude of large conglomerates, which avoid direct exposure in political disputes so as to protect their cloud infrastructures and respective market shares without alienating public opinion.
Concerns related to artificial intelligence, however, are now also affecting institutional levels. To give an example, the New York City Council has scheduled an extraordinary public hearing entirely dedicated to artificial intelligence for October 5, inviting both OpenAI and Anthropic to formally participate, with explicit reference to the recent wave of reports and analyzes that warn of the technology's potential.
