Last week, the top of the Australian Electoral Commission said warned The organization “doesn’t have the legislative means or the inner technical capability to discourage, detect, or adequately take care of false, AI-generated content in regards to the electoral process.”
This remark before a Senate committee on Introduction of artificial intelligence (AI) just isn’t an isolated commentary. The relationship between AI and democracy is the topic of various, increasingly urgent conversations around the globe.
More than 60 countries will go to the ballot box in 2024, in a synchronized “the most important election 12 months in history.” Australia expects elections within the Northern Territory and Queensland this 12 months, with federal elections scheduled for May 2025.
At the identical time, the explosion of generative AI tools for text, images, audio and video is dramatically changing the way in which Australians create and process information. How can we maintain electoral integrity and trust within the age of generative AI?
Deepfakes and disinformation
The most evident risk that AI poses to democracy is synthetic content (or “deepfakes”) that could possibly be used to deceive voters. World Economic Forum survey A survey conducted last 12 months found that experts ranked “misinformation and disinformation” and “social polarization” as the primary and third most serious global risks for the subsequent two years.
These risks are already apparent. In the United States, a political consultant who used the synthetic voice of US President Joe Biden in robocalls faces fines of several million dollars. In India, AI-generated videos have gotten increasingly common on this 12 months’s election campaign.
But deepfakes and misinformation are removed from the one risks. AI also offers latest opportunities. I actually have reported to the Senate Committeeand in a Template My colleagues on the Tech Policy Design Centre and I imagine that a broader national discussion on this issue is important.
A comprehensive view of a healthy democracy
A comprehensive policy will focus not only on deepfakes influencing votes, but on the health of democracy more broadly. Free and fair elections are one feature of democracy (albeit an incredibly essential one), but there are a lot of others.
Informed civic engagement, tolerance and political pluralism are other essential components of a thriving democratic system. The system must also recognize and reply to the needs of voters, and the federal government should be transparent and accountable.
So when we expect in regards to the relationship between AI and democracy, we want to grapple with enduring themes reminiscent of political representation, public interest journalism, media literacy, and social cohesion.
Risks – but additionally opportunities
A balanced policy should recognize that AI technologies pose each opportunities and risks for democracy.
For example, it’s entirely reasonable to fret that informed civic engagement might suffer from the tendency of generative AI models to “hallucinate“ and produce misinformation.
However, the identical technology can even involve more voters within the civic debate: it may well transform complex political concepts into comprehensible content or produce automatic translations into many languages.
Learning from the past
Some elements of the challenges we face aren’t as latest or certain as they appear.
AI, and particularly the boom in generative AI, undoubtedly brings some unprecedented elements to the democratic ecosystem. But there are still lessons we will learn from the past.
Fears about technologies that make latest types of communication accessible to everyone are nothing latest. Efforts to manage and control who can influence public information flows are also nothing latest.
In the fifteenth century the invention of the printing press by Gutenberg stoked fears about what we might now call “fake news.” There are many newer examples, including what we now have learned from the rise of social media (which in turn shapes the story of AI and democracy).
We need to acknowledge the weather of generative AI which can be fundamentally latest. At the identical time, we will search for applicable policy tools and lessons from previous information technology revolutions.
What now for Australia?
Australia is at an interesting inflection point. With a 12 months to go until our next general election, several interrelated policy areas are developing.
The Ministry of Industry is working on a response to last 12 months's Safe and Responsible AI AdviceThis includes considering rules for mandatory watermarking in AI-generated content.
The Communications Department can be revision Proposals for brand spanking new powers for the Australian Communications and Media Authority. These will help combat misinformation and disinformation and should include AI-specific measures.
At the identical time, the Online Safety Act 2021 checkedThis may result in powers to tackle online abuse of public figures, where AI could play a job too.
I propose 4 key actions that the Australian Government should take.
First, it should develop a coordinated national approach to the connection between AI and democracy. My colleagues on the Tech Policy Design Centre have suggested More detailed recommendations on how the event of national technology policies might be coordinated.
Secondly, the federal government should pay special attention to the Dozens of national elections worldwide this 12 months. We can observe the success or failure of various strategies in numerous contexts to learn from the experiences of others.
Third, we will learn from South Korea by requiring politicians to reveal and watermark deepfakes or other AI-generated content utilized in campaign materials. South Korea has completely banned politicians from using AI-generated materials of their campaigns. However, the lower hurdle of requiring Australian politicians to be transparent is more likely to be less controversial and easier to implement.
And fourth, the federal government must make sure that the Australian Electoral Commission and the Australian Communications and Media Authority have the staff and resources they need to deal with latest challenges, including those posed by AI, and equip Australians to take care of a fancy information landscape within the 12 months ahead won’t be a simple one.
These steps must be just the start of a comprehensive, balanced and informed national discussion about how we can assist Australia's democracy thrive within the age of artificial intelligence.
