If There’s a Will, There’s a Workaround
Why This 25-Year-Old is Afraid of the Internet’s Future… And Still has to Work in It.
I’m scared of where the world is heading. There, I said it.
Change is hard for me (hello, autistic brain), and the pace of everything (policy, platforms, AI, etc.) feels like standing on a treadmill that someone else keeps cranking up. Lately, the trigger was Australia’s policy push for age verification (and limits) on social media, search engines, and adult content. I understand the intent: protect children. But I also know the internet’s oldest adage: if there’s a will, there’s a way. And workarounds can shove people - kids and adults - into darker, riskier corners of the web.
Below is me trying to make sense of it all: what’s changing, why it’s terrifying, and where that leaves us as individuals (and as someone who actually works in content and branding… on the very internet I’m side-eyeing).
What’s changing in Australia (and why it’s complicated)
Australia is rolling out age assurance across a wide swath of the internet. From December 2025, age checks begin for adult content via ISPs/search, with a broader rollout in March 2026 that reaches websites, social media, app stores, and even AI tools. Penalties for non-compliance are heavy. The regulator, eSafety, has been laying the groundwork for years, proposing options like credit card checks, photo ID, facial age estimation or digital ID wallets.
Reference: The Guardian | News.com.au | eSafety
On paper, this focuses on harm reduction. In practice, we’ve seen circumvention elsewhere. When the UK enforced age checks for adult sites in 2025, VPN downloads spiked and traffic shifted. Sometimes away from big, regulated platforms to smaller murkier ones. In the US, when Utah required age verification, some major sites simply blocked the state, and users sought alternatives. None of that exactly screams “safe”
Reference: The Verge | WIRED | Kuer | Deseret News
And we’ve been here before: the UK tried a porn age-check plan in 2019 and ultimately dropped it amid privacy and practicality concerns, then revived age-gating under a different law years later. That whiplash matters: it shows how technically hard and socially messy this is.
Reference: WIRED | SCL | Wikipedia
We’re trying to child-proof a hurricane. The likely side effect is not just “less access,” it’s different access through often riskier routes.
Two more accelerants to my anxiety
1. The Speed (and weirdness of AI)
Every week brings a new capability leap. Text-to-video models like OpenAI’s Sora compress what used to be a film crew into a prompt. Multimodal models like GPT-4o talk, see, and respond in real-time. Google’s Gemini 1.5 popularised the idea of reading massive context windows (think entire books, videos, and codebases at once). I can’t keep up and I work in this space.
Reference: Open AI - Sora | Open AI - GPT-4o | Android Headlines
The risks are not theoretical. In 2024, a consultant used an AI-cloned voice of President Biden in robocalls before the New Hampshire primary and was hit with a proposed $6M FCC fine (which was later finalised). That’s not sci-fi; that’s disinfo-at-scale.
It’s wonder and whiplash at the same time. The creative upside is real. So is the possibility of being tricked by something indistinguishable from reality.
2. Over-monitoring by tech giants
We’ve lived through Cambridge Analytica, where Facebook data on tens of millions was siphoned for political profiling. We learned Google could log your location even with “Location History” off (because another setting was still on. Surprise!). Amazon Ring has shared doorbell footage with police in “emergencies” without owner consent, and its law-enforcement integrations keep evolving. TikTok’s in-app browser once included code capable of monitoring keystrokes on third-party sites. Each story is different; the pattern feels the same.
Reference: Wikipedia | ABC | Axios | AP News | The Guardian | Ars Technica | ABC News
Meanwhile, the business model of the web is still personalised ads built on data trails from searches, sites, apps and location. Even as Chrome phases out third-party cookies and pivots to Privacy Sandbox/Topics, the goal remains: show you ads tailored to your behaviour. Regulators are chipping away at the worst abuses (eg. the FTC’s location-data crackdowns), but the ecosystem is sprawling.
Reference: Privacy Sandbox | Google Help | Federal Trade Commussion
I don’t want to live inside a marketing experiment. I just wanted to Google corset sewing patterns and post my work.
A new headache for my industry: GEO (Generative Engine Optimisation)
If you thought SEO was a moving target, welcome to GEO. GEO optimise content so that generative engines (ChatGPT, Gemini, Perplexity, AI Overviews, etc.) cite you or summarise you. The original research formalised GEO and even found content tweaks could raise “visibility” in generative answers by up to 40%. It’s fascinating (and existential) because a model might answer the user directly, reducing clicks to the source. As a branding specialist and content writer, I now have to write for both humans and for models that feed humans. Terrifying, but also (unfortunately) now the job.
Reference: Cornell University | Search Engine Land
Some quick definitions:
Generative engine: A search interface that synthesises answers (not just links) using LLMs
GEO: Strategies (structure, citations, statistics, clarity) that increase the odds your work is cited or excerpted in those answers
I finally understand why our grandparents stared at beige computers like they were alien toasters. I’m only 25 and already nostalgic for last year’s internet!
Why workarounds often make things LESS safe
When access gets gated, determined users often:
Install VPNs or use shade “free” ones (which may log or sell data)
Drift to unregulated mirrors or sketchy aggregators with malvertising
Share content via encrypted groups or side-channels that are harder to moderate
We’ve seen versions of this in the UK (VPN surges and traffic displacement and the US (platform blocks prompting people to go elsewhere). “Safer internet” is the goal, but safety theatre can push risk underground.
Locks without ladders just create better climbers.
So… why am I scared?
Because the rules (and the rule-breakers) are both getting smarter. Because AI is both a paintbrush and a fog machine. Because the businesses that connect us are also the businesses that study us. And because I’m 25, neurodivergent, and trying to grow not only my career but my business inside this roaring machine while wondering whether the ground under my feet is a conveyor belt.