Five AI and Disinformation Stories You Need To Read This Week
We protect everyone from banks to elections from online manipulation
What's happening this week in the world of AI and disinformation? We've curated five stories to keep you up to date. We're Valent, and we identify disinformation threats with our research team and AI tool, Ariadne. Here's what has piqued our interest this week, along with our hot takes:
AI could 'supercharge' election disinformation, US Deputy Attorney General tells the BBC
There seems to be a consensus developing that Generative AI is dangerous because people are more likely to believe something that looks real. But, as Carnegie’s recent Countering Disinformation report notes “people’s willingness to believe false (or true) information is often not primarily driven by the content’s level of realism”. Focusing on restricting Generative AI is likely to have as much impact on disinformation as promoting fact checking - which is to say it might have a bit of impact, but isn’t the whole picture - Amil Khan
Germany's top security official wants easier ways to track right-wing extremist financing
The new powers and capabilities Germany’s interior minister is calling for include an “early recognition unit” to detect disinformation and influence campaigns. The recent announcement by France’s own Viginum that it had uncovered a Europe-wide Russian network shows how vital such government capabilities are. But standing up such a unit comes with legal and institutional challenges; including, figuring out where a unit that monitors foreign and domestic disinformation sits within a traditional government structure, protecting freedom of expression and respecting existing privacy laws - Amil Khan
TikTok to open in-app Election Centers for EU users to tackle disinformation risks
One of TikTok’s efforts to protect election integrity in the EU involves offering users a prompt within the app that takes them to “trusted information”. There are a number of questions with this approach. This kind of intervention requires a social media platform to identify (either directly or via an intermediary) what information on its platform is worthy of a warning, and where it should direct users to in order to protect them. This is a strange approach for a platform to take when the question of what is and isn’t trustworthy is in itself such a contentious issue - Amil Khan
Sora: OpenAI launches tool that instantly creates video from text
The process will be agile i.e. finding the best results and testing different ideas as opposed to manually grinding out the content with a waterfall approach. We've had the same situation for software engineering where we've had code generators available for the last year and it definitely enables! - Fergus McKenzie-Wilson
Global Accounting body publishes report on the Impact of AI on Accounting and Finance Professionals
Recently I saved several hours of my day by asking ChatGPT to write a complex nested excel formula for me. It is becoming clear that in many industries, AI is transforming how we work. The Institute of Management Accounting (IMA) published a report considering the Impact of AI on Accounting and Finance Professionals and key findings are that organizations must “proactively embrace AI-induced changes by fostering human-machine collaboration”. I think at this point it is clear to most businesses that if you are not thinking about how to integrate AI into your business, you will be inefficient compared to your competitors. Some advice offered is to upskill your workforce to create efficiencies in their use of AI tools, and to create a culture of ongoing learning - Kaleem Khan
That's all for now - remember to follow Valent Projects on LinkedIn to stay on top of developments in the AI and disinformation threat space.