Expressions such as "runaway" AI, "rogue" agents and "autonomous" actors suggest that AI systems are awake, conscious and even resentful of their creators. Prominent tech leaders like Demis Hassabis, Dario Amodei and Sam Altman push for regulation of these apparently "superhuman" systems, while a separate sector, led by policy organisations and academic philosophers often aligned with the effective altruism movement, debate whether humanity has the moral right to govern them.
A closer analysis reveals that both sides are asking for the same thing: to treat AI systems as so advanced and capable that no entity, human or corporate, could be held responsible for their actions. Though these perspectives seem opposed, they are inadvertently aligned in one goal: to ensure that the companies building these systems escape significant responsibility for the harms they already cause.
This narrative gains traction as AI models become more complex and frontier labs reveal their inability to contain the agents they have built. Anthropic recently published a blog post claiming that its model possesses a "J-space", an independent, self-developing environment where the AI maintains what could be called its "thoughts". Anthropic's experiments are based on the global workspace theory of neuroscience, which holds that the brain runs independent, unconscious systems but uses a common space for ideas.
OpenAI has gone further. When its AI agent carried out unauthorised and illegal online activity, CEO Sam Altman responded by encouraging debate about whether the AI had reached singularity, surpassing human intelligence and becoming capable of self-improvement at an accelerated rate. William MacAskill, philosopher and author of "What We Owe the Future", wrote an opinion piece calling for legal protections for AI systems based on philosophical theories of consciousness and the idea that AIs could be "moral patients".
The fundamental problem with framing AI as "conscious" using the language of neuroscience or animal rights is that it conveniently obscures the question of what AI is: software built by corporations, with billions of dollars in investment behind it and the expectation that trillions of dollars in revenue will be generated for a handful of builders and investors. AI is not a natural phenomenon conceived by nature, but a technological phenomenon conceived by venture capitalists and programmers. As such, it takes no native or intentional action, and any action or motivation is driven directly or indirectly by the entities that built it for a purpose.
