At the AI Impact Summit 2026 in New Delhi, Indian publishers push back against treating journalism as free training data. They argue that training data from journalism should be paid for and treated as intellectual property, created with investment, infrastructure, and talent. The panel framed AI in News and Journalism Data as twin pillars of responsible progress, urging models to distinguish news from generic data to improve accuracy and reduce hallucinations. LV Navaneeth of The Hindu stated clearly: Journalistic content is not free-floating internet material; it is intellectual property that must be contracted. The session featured Kalli Purie of India Today, Mohit Jain of Bennett, Coleman & Co, Pawan Agarwal of Dainik Bhaskar, Robert Whitehead of INMA, and Tanmay Maheshwari of Amar Ujala, with Ashish Pherwani of EY moderating.
AI in News: The credibility edge for editors
On AI use in the newsroom, publishers stressed that AI in News is a tool, not a replacement for reporters. India Today’s Purie described an AI sandwich: human intent starts the AI process, AI handles what it can, and a human makes the final decision. The Hindu’s Navaneeth echoed this view, saying AI in News should deepen readers’ understanding rather than replace editorial judgment. The Hindu also cited its in-house model trained on its archives, designed to reduce hallucinations by leveraging internal data. Amar Ujala’s Tanmay Maheshwari highlighted multilingual challenges, noting accuracy below 55 percent for many Indic-language models. This Journalism Data-driven approach could help maintain public trust.
Industry observers noted that the value of credible reporting rises when Journalism Data underpins AI tools with verified sources. The panel argued that the digital ecosystem should reward accuracy and accountability, not just speed or novelty. The discussion emphasized that AI in News can augment, not erode, the journalist’s role in curating trustworthy context for readers.
Journalism Data: Training, transparency, and public good
A Journalism Data frame could map training material to sources, ensuring journalists are fairly compensated and credited. Purie emphasized transparency about data sources scraped from publishers’ websites, and argued for reliable labeling so readers can trace AI outputs back to the originals. Several speakers echoed the idea that Journalism Data should be treated as a public good, rewarding stories that deliver social impact rather than chasing clicks. The call was explicit: honor verified content, penalize AI hallucinations, and foster a fair digital marketplace where publishers receive fair value for their material.
GLOBAL scene: The global debate mirrors local concerns. The Digital News Publishers Association has pressed court cases and policymakers to require permission and payment for training data. The Times Group, The Hindu, India Today, Dainik Bhaskar, and Amar Ujala used the summit to discuss revenue models, licensing, and how to measure trust in AI-assisted stories. INMA and other associations emphasized accountability standards for AI-generated content and for platforms distributing it.
Purie also outlined a nine-point plan emphasizing data-source transparency and traceability. She urged clear labeling of AI-generated content so readers can trace outputs to the original story. Several speakers backed recognizing Journalism Data as a public good and asked tech firms to reward stories with social impact, not just virality. Put simply: verify content, penalize hallucinations, and reward credible reporting.
Meanwhile, Whitehead urged policymakers to consider a law requiring paid training of AI models on journalistic content. He pointed to Norway and South Africa as early explorers of a fair digital marketplace where proceeds flow to the creators of trusted material.
Practical steps for AI in News and Journalism Data
- Adopt licensing frameworks that clearly tie usage of Journalism Data to compensation for publishers and authors.
- Institute transparent data-source labeling and traceability so readers can identify the provenance of AI-generated outputs.
- Balance automation with human oversight: AI in News should augment newsroom capabilities while preserving editorial judgment.
- Develop public-good incentives that reward stories with social impact, rather than chasing sensational virality.
For publishers, the practical path includes testing in controlled pilots, building clear attribution norms, and aligning with regulators to create a fair digital marketplace where Journalism Data is valued and protected.
FAQ: AI in News and Journalism Data
- What is AI in News? It refers to using AI tools to assist reporting, editing, fact-checking, and storytelling, while keeping human oversight central.
- Why should Journalism Data be paid for? Because it represents original reporting, investment, and editorial labor that creates value beyond raw data.
- How can readers trust AI-generated content? By clear labeling, source tracing to verified articles, and accountable editors who review AI outputs.
- Will this stifle innovation? Proper licensing and fair compensation can incentivize high-quality work while enabling responsible AI development.
Readers are encouraged to share perspectives in the comments as the debate evolves across 2026 and beyond.
Conclusion: AI can amplify journalism if editors stay in control, and if platforms and lawmakers work with publishers to ensure fair compensation and transparent sourcing. The summit left attendees hopeful that tools can assist, not replace, the newsroom’s core mission.
Original reporting and insights from The Hindu and other DNPA members deserve thanks. A sincere thank you to The Hindu for the original reporting; see the coverage at The Hindu and related outlets.
Tell us what you think in the comments below as we continue this conversation about AI in News and Journalism Data in 2026 and beyond.
Global context and next steps
Beyond India, several regulators and industry groups are examining the economics of training data. The idea is to ensure that paid, ethical sources support AI systems in a way that sustains quality journalism and public trust. As platforms experiment with new monetization models, publishers emphasize that credible, well-labeled content should be the premium input that powers responsible AI.
References
Further reading
- OECD AI Principles
- The Verge: What counts as fair training data for AI?
- Nature: The ethics of AI training data