The conversation about the “Dead Internet” has ceased to be a niche issue and has become a central topic of technological debate. The trigger was a recent comment by Sam altman, CEO of OpenAI, who has questioned how much online activity is still truly human.
Altman slipped into X that he never paid much attention to this idea, but that today he detects many accounts managed by linguistic models. Having the head of one of the companies that has most popularized generative AI say this has set off alarm bells, and in the process has once again brought into focus a hypothesis that is as uncomfortable as it is difficult to prove precisely.
What the theory proposes and where it comes from
The so-called theory of the “Dead Internet” He argues that a very significant portion of what we see—posts, comments, reviews, or even “people”—would not be produced by real users, but by bots and algorithmsIn this vision, the web would be a setting where a few humans coexist with a swarm of simulations.
Its origin is located between mid-2010s and 2017, when a shift began to be noticed in forums and communities: less spontaneity and more repetitive, homogeneous content optimized to attract clicks. The rise of generative AI in recent years has given this suspicion a veneer of plausibility it didn't have before.
What Sam Altman Said and Why It Matters
In his message, Altman came to recognize that he increasingly perceives more profiles in X managed by LLMThese powerful language models are the foundation of ChatGPT and similar tools like Claude, which can compose compelling text in seconds and hold conversations that sound human.
Appreciation does not come in a vacuum: the release of ChatGPT by the end of 2022 accelerated the arrival of related systems and made mass generation of posts, replies, and pieces of content accessible to anyone. This has facilitated both legitimate uses and abusive practices, from commercial automation to spam and misinformation.
That the CEO of OpenAI verbalizes this concern amplifies the debate because his company has been key lever of the phenomenon. The message also fits with the growing problems of platforms like X to contain bots and account farms, a challenge that has proven more resilient than expected.
Cross reactions and the accusation of paradox
The responses were not long in coming: many users reminded Altman that the expansion of generative AI has multiplied the volume of synthetic content on social media, blogs, and in the media. For its critics, warning about the problem now is like pointing out the smoke after having fanned the flames.
Others interpreted his comment as part of a broader strategy: if one assumes that the Internet is saturated with automation, it would make sense to push systems that distinguish between humans from botsThis reading connects directly with another of the initiatives led by Altman.
Is the network "dead" or just saturated? Data and nuances
Cybersecurity consultants have been warning for some time that a very high fraction of the global traffic It comes from automated systems. Not all of this flow is malicious: there are also search engine crawlers and monitoring tools. Even so, the line between authenticity and synthetics has become more blurred.
The visible effect is a web where fabricated reviews, generic comments and cloned posts proliferate, while powerful platforms are filled with AI-produced videos, images, and texts. Cases like YouTube channels with invented stories, strange images on social media or controversies over the use of virtual models in advertising fuel this feeling.
It is important not to fall into sensationalism: the Internet continues to host living communities, the online creativity and the conversations of millions of people. The metaphor isn't that of a cemetery, but rather a landscape where the human and the automated compete for attention, often with the latter having an advantage.
Digital identity and the World Network (formerly Worldcoin) initiative
In parallel, Altman promotes WorldNetwork (formerly Worldcoin), a project that aims to offer “proof of humanity” through the iris scanThe promise: that any user can prove they're a real person in online environments increasingly populated by automated systems.
The idea has generated a strong controversy, especially in Europe, due to the treatment of biometric dataAuthorities and experts have urged caution, highlighting risks to security, governance, and access to extremely sensitive information.
The director herself AEPD, Mar Spain, went so far as to warn of the implications of handing over such sensitive data in exchange for financial incentives, insisting that this information could have short- and long-term consequences. Following this pressure, the project has focused on strengthening Privacy and transparency, according to those responsible.
Its proponents argue that if automation continues to advance, robust verification of the human condition will be essential; its detractors see it as a disproportionate solution which opens the door to new abuses. The debate, for the time being, remains open.
With all this context, Altman's intervention has worked as a catalyst: it has united in a single conversation the rise of the LLM, the difficulty in tackling bots on massive platforms, the fragility of digital trust and the search for methods - controversial or not - to test who is who On the web. The "dead internet" may not be a literal diagnosis, but it is a warning about a reality: the balance between authenticity and automation has shifted, and deciding what we want the web to look like in the coming years can no longer be delayed.
