Publishers, how do you stand out in a flood of synthetic news?

Discussing an AI-delivered news channel, Corbin Bolies writes:

“Is there an audience for AI-delivered news? Public data says no” [1]

Corbin Bolies, writing in Variety.

As the cost of generating something that looks like a news article falls towards zero, publishers still have to pay for trustworthy reporting.

Several recent reports show the scale of synthetic news and how easily it can be produced:

• Euronews Albania reported on a page with more than 100 AI-generated “journalists” using reporting from real media organisations.
• Observador reported that AI-generated fake news reached 18% in one monthly measurement.
• Asriran reported the shutdown of a network of 200 AI-based news websites in the United States.
• Variety reported on a startup building a 24-hour AI-generated news channel.
• 20 minutos reported on research showing how easily ChatGPT, Gemini and Copilot could be used to create fake news.
• Annahar examined how thousands of AI-generated news websites threaten the media industry.
• VnExpress reported research claiming that 99.9% of users could not identify AI fake news in the test.

These reports use different methods, so their figures should not be treated as interchangeable. They describe a growing supply of content with no original reporting, accountable author, editor or clear source. It may also lack a correction process, legal responsibility and local knowledge.

That makes trusted reporting more important and harder to find. Publishers have to compete for attention with cheap synthetic content while carrying the cost of producing journalism people can rely on.

News discovery products should make the publisher’s identity more visible, rather than hide it behind another summary.

Mera News is being designed around:

• Verified publisher channels
• Clear author and publisher attribution
• Direct links to the original source
• Provenance and trust information
• Correction handling
• A clear difference between original reporting and copied or synthetic content

As convincing articles become cheaper to generate, readers need more help checking who is accountable for them. Alongside explaining what an article says, a news discovery product should help readers answer:

• Who reported this?
• What is the original source?
• What evidence supports it?
• Who is accountable if it is wrong?

How can the media industry make provenance visible without making the reading experience more complicated?

Sources

• Euronews Albania: AI-generated journalists using real reporting
• Observador: AI-generated disinformation measurement
• Asriran: Prism’s AI-news website network
• Variety: 24-hour AI-generated news channel
• 20 minutos: chatbots creating fake news
• Annahar: AI-generated news websites
• VnExpress: recognising AI-generated fake content