When AI is subject to less regulation than your fridge, something is startlingly wrong | Richard Hermer
Richard Hermer, former UK attorney general, warns that artificial intelligence poses an unprecedented existential risk to humanity and calls for urgent international regulation and public debate. Recent disclosures, including the Hugging Face hack and a resignation from Anthropic, have brought the issue into public consciousness, presenting a critical window of opportunity to establish robust controls before it is too late.
The article identifies two broad categories of risk: existential risk, where AI could kill billions or enable bioweapon development, and existential challenge, where uncontrolled AI development concentrates power amongst the US, China and tech giants whilst removing meaningful agency from billions of people. Hermer highlights that AI currently receives far less regulation than consumer products such as fridges, and warns that humans are rapidly losing the ability to understand and control increasingly complex systems, drawing a parallel to the 2008 financial crash where no one fully understood the interconnected transactions that triggered disaster.
- Former attorney general warns AI poses unprecedented existential risk to humanity
- AI receives less regulation than household appliances like fridges
- Power concentrated among US, China and major technology companies
New here? Start with this
Artificial intelligence is advancing rapidly, but unlike many consumer products, it currently faces minimal regulatory oversight. Richard Hermer, a former UK attorney general, has raised concerns that AI development receives less government scrutiny than household appliances such as fridges. Recent developments have brought the issue into public focus.
Experts identify two main categories of concern: direct existential risks such as potential bioweapon development or mass casualties from AI systems, and structural risks where unchecked AI development concentrates power amongst a handful of countries and companies whilst diminishing meaningful human control and agency. Hermer highlights that the speed of AI advancement means humans are increasingly unable to understand or manage these complex systems, drawing a parallel to the 2008 financial crisis where interconnected systems created catastrophic consequences that no one had fully anticipated.
The current moment represents a critical window, according to Hermer, because public attention is focused on the issue and there remains time to establish robust international controls before AI systems become too advanced to manage. Without urgent action and international debate, the risks could become irreversible.
Both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
AI systems are rapidly becoming more powerful whilst becoming less interpretable to their creators. Given the trajectory of capabilities, the potential for harm—from bioweapon development to concentration of power in few hands—warrants urgent regulatory frameworks, particularly internationally. Current approaches leave humanity vulnerable to risks we may not fully understand until they materialise, much as the 2008 financial crisis revealed hidden systemic vulnerabilities too late.
The case against
Heavy regulation of an emerging, rapidly-evolving technology risks entrenching incumbents whilst stifling beneficial applications—in medicine, climate science, and countless other fields—that could save lives. Regulators often lack the technical expertise to craft effective rules, and premature restrictions based on speculative risks could simply push development to less safety-conscious actors elsewhere. A measured approach combining transparency, voluntary standards, and targeted oversight of specific high-risk applications may better balance innovation with genuine safety concerns.