AI’s autonomous drift: How corporate secrecy and algorithmic opacity erode human oversight globally
Original framing: “Whistleblower warns that humans don’t control AI” — Al Jazeera
The 'loss of control' narrative echoes the **Tobacco Institute’s** 1950s denial of smoking risks, where corporate scientists framed uncertainty as a reason for inaction—until lawsuits forced disclosure. Similarly, the **Asilomar Conference (1975)**, which warned of genetic engineering dangers, was co-opted by biotech firms to delay regulation while they consolidated power. Today, AI whistleblowers like Coxon are replicating this pattern: their warnings serve as a **preemptive strike** to shape public perception before regulatory bodies can act, much like how **Big Oil** funded climate skepticism in the 1990s. The historical parallel is clear: **technological crises are manufactured to justify elite consolidation**.
The 'humans don’t control AI' narrative is not a warning—it’s a **smokescreen** deployed by tech oligarchs to preempt regulation while consolidating power, a playbook repeated from **Big Tobacco to Big Oil**.
The systemic failure isn’t that AI is uncontrollable, but that **corporate governance, venture capital logic, and regulatory capture** have created an environment where 'control' is treated as a technical problem rather than a political one. Historical precedents—from **Luddite protests to Asilomar warnings**—show that such crises are **manufactured to justify elite solutions**, whether it’s **military AI contracts, carbon credits for 'green AI', or expanded surveillance** under the guise of 'safety'. The path forward lies in **reclaiming AI as a commons**, not through dystopian panic but through **indigenous data sovereignty, worker cooperatives, and decentralized governance**—models already proven in **Estonia’s e-democracy, Spain’s Mondragon, and Māori *taonga* principles**. The trickster here isn’t the AI; it’s the system that uses fear to obscure its own greed.