Rather a lot happened in 2025. IBM exceeded $50 billion in spend on the acquisition of companies built around open source software. AWS, Microsoft and Cloudflare proved, again, that 100% uptime is a promise only liars make. And there was suddenly a lotmoreAI around.
But in the sweep of general history, I reckon 2025 will be remembered as the year the bans started.
Most recently – and to the mostpublicattention – the ban of under-16s on social media by Australia.
The thing about protecting children is that you can't not. We all know parts of the internet are bad for children, the only argument against age-gating has been that it won't work. Now Australia is doing it.
Either others will follow that example, because they can't not, or Australia will fail in a fashion spectacular enough to push the next attempt out by 10 or 15 years.
There was less notice of Russia's attempts to ban WhatsApp, preferably in favour of its alternative Max, in a combination of reasserting local culture and as a national security measure. Both those issues resonate for countries as diverse as France and South Africa.
Russia does crazy stuff (such as invading its neighbours), and is not the kind of country you emulate directly. Make no mistake, though, its efforts to de-Facebook will be watched almost as closely as Australia's age restrictions, and may well be emulated in 2026.
Ransomware
As a not-too-distant second, 2025 will be remembered for ransomware. We can hope as a low point, as the last year in which ransomware gangs make off with fortunes while costing governments and companies vast sums.
Realistically, of course, we're probably not at peak yet. Policymakers can't even agree yet on a robust response to ransom attacks, and it feels like the average CVE rating just keeps climbing. Meanwhile, while cloud LLMs halt no-code attacks, spinning up a local evil-AI instance just gets easier.
The only argument against age-gating has been that it won't work. Now Australia is doing it.
Take it in sum – no change in incentives, lots of attack vectors, and systems that democratise attacks – and there is, in fact, every chance that 2025 will be remembered as the start of the dark age of ransom attacks.
Sovereignty
The third major theme I'd expect to see in history books in reference to tech in 2025 is sovereignty, with a photo of Donald Trump in the footnotes.
In November 2024, Europe was not entirely thrilled at having all its data on American clouds, but hey, that's where the internet was.
A year later, bringing their data home is a rampant obsession in European capitals, and in not a few other places too.
That kind of fervour is catching, but the pressure in the private sector seems to be a little different. Companies started looking at, or even moving to, hybrid and on-prem for a range of reasons: to get more AI-ready, for more predictable future costs, for fear of vendor lock-in, and yes, for fear of regulators.
The ultimate result is the same, though, in that everyone suddenly wanted to bring their data home, in the biggest reversal in sentiment since the first time somebody scribbled a cloud shape on a whiteboard in a sales meeting.
Then again, it is always possible that we are in the middle of the biggest bubble in the history of the world, and it goes pop in 2026. In which case, 2025 will only ever be remembered as the highwater mark of a mad obsession with throwing money at AI.
I'm rather hoping the bans win.
Share