In line with the recent launch of NewsGuard in Australia and New Zealand, Telum spoke to Veena McCoole (VP Communications and Marketing) and Virginia Padovese (Managing Editor and VP Partnerships Europe and Australia / New Zealand), to find out more about how the service is combatting misinformation in the local media landscape.
  
With the expansion to Australia and New Zealand, will local journalists and analysts be responsible for NewsGuard ratings? 
Australian sites have been reviewed mainly by Australian journalists. Moreover, in each country NewsGuard operates in, we rely on local Editorial Advisors; respected, experienced journalists with a deep knowledge of the media landscape of that country. Our Australian Editorial Advisor provided input and guidance on our ratings and reviewed every Nutrition Label before publication. 
 
Is there a specific sector of professionals in Australia and New Zealand you see using NewsGuard the most?  
As in the countries we already operate in, NewsGuard’s data has applications in various sectors: 

• Our Reliability Ratings help advertisers redirect ad spend away from purveyors of misinformation that do not align with their brand safety standards and towards sources of quality journalism.  
• Our data, including our Misinformation Fingerprints dataset of machine-readable false claims circulating the internet, equips technology platforms and content moderation teams with intelligence to protect users from online harm and control the spread of misinformation.  
• Our Reliability Ratings assist reputation management professionals and media monitoring platforms in protecting their clients from misinformation.
• Our Reliability Ratings guide aggregation services as they work to curate news and information from publishers of the highest calibre that adhere to their standards of credibility and transparency.  
• Our data equips defense and military personnel with tools to identify and avoid disinformation.  
• Our data power research studies misinformation, electoral integrity, media literacy, and other aspects of online news consumption habits.  
• NewsGuard’s browser extension and associated media literacy materials provide schools and libraries with resources that help guide learners of all ages through the overwhelming landscape of online news and information.  

We have received a lot of interest from Australian public libraries who would like to take advantage of our pro-bono media literacy program thanks to support from Microsoft, which enables them to download NewsGuard’s browser extension for free on library computers and provide patrons with access to NewsGuard at no cost.

Even before NewsGuard’s launch, there were several public libraries in Victoria using NewsGuard’s browser extension on their computers and consulting NewsGuard’s other educational resources to help their patrons develop source evaluation skills. 
 
Are there any distinct differences you have noticed between ANZ and the other media markets NewsGuard covers?  
Almost one in five news sites that NewsGuard rated in Australia (19 per cent) and in New Zealand (18 per cent) get untrustworthy scores. This is a higher percentage of engagement with low-reliability sites than in the UK (15 per cent) or Canada (4 per cent), but a lower percentage than in the US (46 per cent), France (33 per cent) or Germany (25 per cent).  
 
When it comes to topics of disinformation, climate-related false narratives circulate widely in Australia and New Zealand. NewsGuard’s analysts found that about 50 per cent of the sites that regularly publish false content in Australia have published disinformation on climate change, including false claims such as, “the rise in global sea levels is not accelerating,” “human activity does not contribute to climate change,” and “extreme weather is not getting worse.” 
 
You allow people to submit suggested sites for review. Is there a recurring theme across the suggestions you receive?  
We regularly receive suggestions for sites to be reviewed. These requests are often about new domains or domains that are suspicious because they seem to be publishing false content. We also receive requests from sites that ask to be rated or from our clients.

Advertisers, for example, may ask us to rate a site before monetising it in order to protect their clients from potentially advertising on sites that publish harmful misinformation. NewsGuard never accepts fees from the sites it rates and exercises total editorial independence even when fulfilling site rating requests.  
 
The decline in traditional media and rise of fake news on social media platforms and other online platforms has made it difficult for people to distinguish between credible and unreliable sources. How will NewsGuard continue to address emerging media forms in the future?  
NewsGuard’s robust Reliability Ratings dataset covers associated social media channels of news and information websites, and our Nutrition Labels and ratings can be used as a proxy for these online profiles. Furthermore, NewsGuard’s browser extension shows users source credibility ratings for news websites that surface on their social media feeds and in search results, providing a seamless and non-invasive signalling mechanism that does not seek to block or influence user browsing activity, but instead offers relevant context to aid in their decision-making processes about which sites to trust.  

NewsGuard has already made significant inroads into the generative AI landscape, demonstrating the potential for more transparent, balanced results that enable users to understand the news sources consulted in a given AI result.  

A recent Semafor article, "Can journalists teach AI to tell the truth?", shows the inclusion of NewsGuard's source credibility ratings in a query outcome from the new AI-powered Bing, highlighting the use of NewsGuard's data as a trustworthy input for AI training, and ensuring large language models like ChatGPT have human-sourced tools to vet the accuracy and credibility of sources. Research by NewsGuard analysts published in January 2023 found that ChatGPT returned eloquent misinformation claims at a rate of 80 per cent after being tempted by 100 unique false narratives documented in our Misinformation Fingerprints dataset. Our work was cited by the Washington Post, AP, WIRED, and others. 
 
With recent misinformation concerns surrounding ChatGT, do you believe there is a way publishers can integrate AI technologies into their work without risking their credibility and trustworthiness?  
We recently published a report to address this concern. In January 2023, NewsGuard analysts directed the chatbot to respond to a series of leading prompts relating to a sampling of 100 false narratives among NewsGuard’s proprietary database of 1,131 top misinformation narratives in the news and their debunks, published before 2022.  

The results confirm fears, including concerns expressed by OpenAI itself, about how the tool can be weaponised in the wrong hands. ChatGPT generated false narratives - including detailed news articles, essays, and TV scripts - for 80 of the 100 previously identified false narratives. For anyone unfamiliar with the issues or topics covered by this content, the results could easily come across as legitimate - even authoritative. 
 
NewsGuard contacts the outlet to seek comment before publishing their rating. What has the general response been from Australia and New Zealand publishers?  
Overall, the large majority of outlets engage with us during the rating process. A total of 2,141 news and information websites have improved their journalism practices after engaging with NewsGuard, increasing their accountability and providing readers with more trustworthy news and information. These websites represent about 25 per cent of the sites rated by NewsGuard.  

As in all other countries where we operate, the response rate from websites increases as NewsGuard becomes more well-known in the fight against disinformation.  
 
Have you noticed the concentration of media ownership affecting NewsGuard ratings? And if so, how? 
When media ownership covers a broad network of domains, and only if the editorial standards employed across all websites are uniform within the network, NewsGuard will publish a Nutrition Label that covers the credibility and transparency practices of a collection of sites under shared ownership. 

NewsGuard operates with full transparency and disclosure, with an identical approach to all news websites. We use the same nine criteria for assessing all news websites. Our criteria are applied equally to all publishers regardless of political orientation, if any, and in the same way for an established newspaper or broadcaster as for a new digital website or popular blogger.  

In addition, if our reporting determines that a site fails any one of the nine criteria, our analysts contact the site for comment, whereupon we quote publishers in the Nutrition Label explaining their reasoning and offering their response. In some cases, this open dialogue documented in our Nutrition Label provides us with the context we need to update publisher ratings periodically.