182 days to save the world from AI

Revived legislative vigour to take on Big Tech suggests politicians are waking up to the need for speed while they can still make a difference.

by Rich Fairbairn

Published: 15:26, 03 July 2025

A week is a long time in AI. The next six months might seem like a lifetime.

In the last three days, two things will have swung the mood inside Big AI boardrooms, from cocky to rocky. For how long remains to be seen, but recent events have given some hope to those who think the overnight incumbents of the emerging sector are already Too Big to Slow Down.

First up, US Senators whipped the rug out from under the AI party when they voted 99-1 to reject a proposed 10-year ban on state-level regulation of AI - meaning individual states can now apply their own laws to AI companies and the use of the tech.

Such local meddling was fiercely opposed by those at the top of the AI food chain who claimed that allowing states to legislate things like kids’ online safety was too detrimental to their share options to be allowed. Also, something something something China.

This is a big deal. If there were no meaningful federal regs in place, a ban on state regs seemed to assure almost unlimited power to Darth Altman et al.

While the subsequent state regs will no doubt be a mix of the mostly sensible and occasionally inane, it does reset the clock and encourage federal legislators to craft coherent rules which actually have some teeth, versus the current “We’ve tried nothing and are all out of ideas!” non-approach.

The Trump White House has shown it is able to move with startling speed and take the Senate with it, and the next few months are being seen as critical in setting the tone for how AI firms are brought to heel, or not.

Denmark hits the ground running

On the same day, Denmark exploded out of the blocks in the first morning of its 6-month EU Presidency and gave some insight into what an ‘AI vs EU’ landscape might look like for years to come, by announcing it was suing OpenAI for mass theft of the nation’s copyright.

Karen Ronde, CEO of the Danish media copyright body DPCMO, announced it in a way that removed any thought that the timing of the case being announced was coincidental: “Denmark takes over the EU Presidency and a united Danish news industry will sue OpenAI.”

Ronde, an ex-MP, said, “Big tech must respect national laws,” and that the case was about fair competition, preventing market dominance, fostering innovation, and denying “any company the power to undermine rivals, exploit content creators, or distort democracy.”

While the DPCMO is technically an independent non-profit, it speaks with the authority of the state: it was established by Government act in 2021 to negotiate copyright matters on behalf of all Danish media, after ministers saw how countries like Australia, Canada, France and the UK forced Google and Facebook to have to deal with local media, and what happened where no such regulatory frameworks existed.

With a mandate from the state, even though DPCMO has only been in existence for four years it s taken seriously and has been pretty successful in a short space of time, with tech firms including Google doing cash-for-content deals.

No-one should be surprised at the turn of events. Ronde made clear what the organisation’s objectives were in a widely published op-ed back in February, whose opening line gave more than a hint at the future: “Curbing Big Tech’s power over media outlets requires publishers, journalists, and photographers to speak with one voice. While the Danish news industry has led the way with a collective management organization, this effort would be bolstered by measures to improve enforcement of the European Union’s copyright law.”

And it tends not to prevaricate. When DPCMO failed to reach a deal with Apple, it simply reported the matter as theft to the police; Apple eventually removed the contested product in Denmark rather than set a precedent of doing a deal.

Basically, the Danes are in a hurry and believe the next six months are critical. As President of the EU it cannot ramrod laws into effect, but it can set a tone which it believes will inspire the others and it on a mission to do so before the year is out, with 182 days to go.

In fact just days before the DPCMO action, the Danish Government revealed world-first plans to give individual Danish citizens anti-deepfake rights over things like appearance, voice, and likeness, while still trying to offer space for comedians to do satirical impressions of techbros and politicians.

The varying pace of change

Meanwhile, as Denmark signals it wants to move at speed and set a pathway which other EU nations can adopt, the bloc itself is wobbling on enforcing its own existing AI laws – not through lack of will but because it rolled them out so fast that plenty of people aren’t really sure what they are or how to comply with them.

Calls are now growing for a slowdown or pause on implementation timelines of the existing EU AI Act, which is due to become enforceable from next month, on the basis that it is too restrictive on homegrown AI firms, and not detailed enough for big users of AI to know where the lines of responsibility are.

The overlap between laws designed to favour the growth of AI, and laws which give powers to individuals against AI, are the next stage of the ongoing battleground of Big Tech vs Governments.

Media and media people will get caught up in it too, with stakes in both sides of the argument.

For those who are convinced the AI boats have sailed, it looks like the regulators have just hopped in a speed boat.