I am making this submission to the Select Committee’s enquiry both as the former MP for Weston-super-Mare 2005-19 who campaigned for stronger measures to address misinformation and disinformation, and who spoke and tabled amendments about the issue while in Parliament. I am also Founder and Director of the Centre for Small-State Conservatives (you can find more details if needed at Centre For Small State Conservatives – Bigger citizens need smaller Governments)
Social media firms’ algorithms are core parts of their business models, which encourage their users to engage more frequently and deeply in each platform’s service by spreading posts which are already attracting other people’s attention, and by sending users new posts about things they’ve shown interest in before. But this brings 2 problems with it:
a) It increases opportunities for bad actors deliberately or recklessly to post attention-grabbing but inaccurate or one-sided information, which make it harder to trust what we see or hear online, and weakens our democracy.
b) It creates ‘radicalisation rabbit-holes’ where users looking at topics like Islamic jihad, political alt-right or hard left, incels, vaccine effectiveness or chemtrails are shown more & more posts which are increasingly one-sided and irrational. This ultimately leads users into intolerant online echo-chambers where they are never presented with the other side of an argument, where dissenting or alternative views are shouted down, and where polarisation, division & potentially-violent extremism flourish.
Q2a What role do generative artificial intelligence (AI) and large language models (LLMs) play in the creation and spread of misinformation, disinformation and harmful content?
AI and LLMs create an arms race where increasingly-capable technology allows the creators of misinformation and disinformation to produce posts faster, more cheaply and make fakes harder to spot. But AI and LLMs also give social media platforms stronger capabilities to identify and react to misinformation and disinformation faster, more cheaply and more accurately too.
This arms race means that regulators and policymakers should be sceptical of arguments from social media platforms that measures to combat misinformation and disinformation are too difficult. The frontier of what is possible and affordable is advancing rapidly and continuously, so providing regulators and policymakers focus on specifying the outcomes which must be achieved, but leave it to the platforms to use the latest and most efficient or effective technologies to deliver them, then the arms race can be harnessed to contribute to the greater good.
Q4 & Q4a How effective is the UK's regulatory and legislative framework on tackling these issues? How effective will the Online Safety Act be in combatting harmful social media content?
Our existing regulatory and legislative framework does not yet address the issues of misinformation and disinformation properly. The Online Safety Act does many important and worthwhile things (particularly in protecting children online) and was originally supposed to include measures to deal with these problems too, but they were dropped at an early stage of its development. It remains one of the biggest and most important gaps in our country’s digital and online protections.
Q4b & Q4c What more should be done to combat potentially harmful social media and AI content? What role do Ofcom and the National Security Online Information Team play in preventing the spread of harmful and false content online
There are 2 changes which would solve many of the problems created by misinformation and disinformation:
a) We should require the major social media platforms to give their users new tools (for example as part of their normal account settings) to decide whether and how they want posts which are factually inaccurate or made up to appear in their timelines or not. These tools should use existing proven, independent journalistic and content-provenance standards for assessing factual accuracy, to prevent meddling by vested political, commercial, religious, criminal or foreign interests.
b) We should extend Ofcom’s proven and longstanding ‘Duty of Balance & Undue Prominence’ which has successfully prevented ‘radicalisation rabbit-holes’ in UK TV & Radio broadcast media for decades, to apply to social media algorithms so they can’t send users more and more extreme posts about a topic without including the other side of an argument too.
These 2 changes will:
I hope this submission helps the Committee’s work. I will be happy to answer any further questions you may have, which arise from this submission.
27 November 2024
High-trust societies are richer and work better, because most things happen faster and more cheaply without needing lawyers or police. Our rivals would like nothing better than for us to distrust each other and start fighting amongst ourselves.
In a democracy that prizes free speech, seeing & hearing the other side of an argument and having to explain why and how you disagree is better than just shouting someone down.
A lie is halfway around the world before the truth has got its pants on [Winston Churchill]
The opposite of knowledge is not ignorance, but deceit and fraud [Jean Baudrillard]
I'm not upset that you lied to me, I'm upset that from now on I can't believe you. [Friedrich Nietzsche]
Far from it. We’re using independent, international standards to give citizens more control over what they see on social media, and how they see it.
The only people who will suffer from these changes are the ones who want to weaken our society and our democracy by telling lies without anyone caring enough to stop them.
Yes they should. But if everybody disagrees with everybody else about everything, you haven’t got a society in the first place. Successful, peaceful, prosperous countries and communities need enough shared beliefs and values for everyone to get along, so we should treat deliberate attempts to undermine or weaken them very seriously indeed.