AI in Media: Schibsted and Reuters offer key insights on adoption and transparency

We used Chat GPT to synthesise these LinkedIn posts from Jane Barrett, Head of AI Strategy at Reuters, and Schibstedmedia group.

These insights provide a comprehensive look at both the importance of transparency in AI use within media and practical steps for organisations beginning their AI journey.

Schibsted Media's AI Transparency Recommendations
Earlier this year, Schibsted Media'sAgnes Stenbom, alongside senior leaders from 13 major Swedish media companies, participated in a collaborative initiative aimed at improving transparency regarding AI usage in editorial content.

Facilitated by the Nordic AI Journalism Network under the umbrella of Utgivarna (the Swedish Publishers' Organisation), the group delivered seven specific recommendations on when and how AI should be disclosed to media consumers:

• AI that has 'significant journalistic impact' must be disclosed to audiences.
• Internal AI tools used behind the scenes without affecting content directly do not require transparency.
• Transparency in AI usage should be treated as an iterative process as technology evolves.
• The media industry should be specific about the type of AI tools being utilised.
• Information on AI usage should be shared in connection with the content being consumed.
• There is a need to harmonise the language used across the media industry regarding generative AI.
• Media companies are advised to avoid visual labels or icons indicating AI usage in editorial content, to prevent trivialization.
• Schibsted Media is committed to responsible AI usage in its products and processes and is actively exploring opportunities to implement AI in a transparent manner.

Jane Barrett's Advice for AI Adoption

For those wondering how to start with AI implementation, Jane Barrett shared her three key recommendations.

Her advice is aimed at companies and teams still in the early stages of integrating AI into their operations:

Barrett stresses that organisations should not delay AI adoption while waiting for perfect solutions, but should instead focus on gradual integration as tools develop.

• Start with any team available, whether they are formally involved with AI or early adopters. Small projects can provide valuable learning experiences about what works and potential pitfalls.
• Gather use cases and prioritise projects that align with the organisation's overall strategy. This approach helps to identify where AI can make the most impact and accelerates its adoption.
• Avoid striving for perfection in AI projects. Generative AI may not always perform tasks flawlessly, so it is important to integrate AI tools with human expertise to create a balanced, responsible implementation.
  

Barrett's post was a summary of her chat to Omar Mehtababout what Reutersare doing with AI and GenAI in the newsroom. Listen to the full episode on:

Spotify:https://open.spotify.com/show/3BnbYyRoElYQRCf1jsklts?si=03241a882bef4bf9&nd=1&dlsi=c3a1673bb3174d61

Apple:https://podcasts.apple.com/gb/podcast/ibc365-accelerating-innovation/id1760009999

WAN-IFRA Staff

[email protected]