Katherine Dee’s Spectator essay imagines a future in which cyberattacks linked to AI make the public internet so risky that people retreat from it. Her provocative twist is that she finds the prospect oddly comforting, even as she acknowledges the disruption and danger it would bring.
Watch Desk analysis
What happened
In an essay published on 8 November, Dee pictures a sustained cyberattack prompting organisations and ordinary people to leave the public-facing internet. She also imagines governments restricting domestic access while essential services remain online. These are scenarios in her essay, not reports of an attack or a proposed government plan.
Dee connects that imagined retreat to AI-enabled cyberattacks, deepfake fraud and the prospect of AI replacing customer-service workers. She also points to Block’s newly unveiled voice app, Dessa, as a potential source of job displacement and fraud. The essay is an argument about what those risks make her feel, not evidence that any of those outcomes is imminent.
Why it matters
The internet has become so woven into ordinary life that a thought experiment about abandoning it quickly becomes a question about what people would lose: banking, shopping, streaming and routine contact. Dee’s essay turns fear of AI-enabled harm into a case for imagining less connectivity, rather than simply better protection.
That is a striking position, but it leaves the hard question hanging. The services people might want to escape are also the ones they rely on, and disruption would not be evenly shared. A retreat from the public internet could mean more than fewer feeds and a pleasant reunion with the telephone.
Our read
Dee makes a memorable case for taking the social costs of cyber risk seriously, but the essay’s imagined internet blackout is not a practical answer to those risks. Its value is as a provocation: it asks whether a more connected world still feels worth having when its weak points are so consequential.
What to watch
- Whether fears about AI-enabled cyberattacks lead to concrete proposals for restricting public internet access.
- How organisations balance online access with the resilience of essential services.
- Whether AI voice tools bring useful services without worsening fraud or job displacement.
Discussion spark: If the public internet became much riskier, would limiting access be an acceptable trade-off for safety, or would it put too much of everyday life out of reach?
Sources and evidence
- What if the internet stopped? – The Spectator (8 November 2026, 13:48 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.