2 Sources
[1]
Meta apologises for inaccurate Kannada translation, claims to have fixed issue
Meta has apologised for errors in its Kannada translation feature. Karnataka Chief Minister Siddaramaiah raised concerns about inaccurate translations on Facebook and Instagram. These errors distorted facts and misled users. The Chief Minister's office formally requested Meta to correct the issue. Meta stated that they have fixed the problem. Social media major Meta on Thursday apologised for inaccurate Kannada translation and claimed to have fixed the issue, after Karnataka Chief Minister Siddaramaiah raised concerns over faulty auto-translation of Kannada content on its platforms. Raising concerns over the faulty auto-translation of Kannada content on Meta platforms such as Facebook and Instagram, Siddaramaiah said that it was leading to the distortion of facts and misleading users. "We fixed an issue that briefly caused this inaccurate Kannada translation. We apologize that this happened," a Meta spokesperson said. Siddaramaiah said social media platforms must act responsibly, especially when dealing with official communication, and cautioned citizens to be aware that translations displayed on these platforms are often inaccurate. "Faulty auto-translation of Kannada content on Meta platforms is distorting facts and misleading users. This is especially dangerous when it comes to official communication. My media advisor has formally written to Meta urging immediate correction," Siddaramaiah posted on 'X'. "Social media platforms must act responsibly. I caution citizens to be aware that translations shown are often inaccurate. Such negligence by tech giants can harm public understanding and trust," he added. In an email sent to Meta on July 16, the chief minister's media advisor, K V Prabhakar, raised serious concerns on behalf of the CM regarding the automatic translation of Kannada content on Facebook and Instagram. The Karnataka government has asked Meta to temporarily suspend the auto-translation feature for Kannada content until translation accuracy is reliably improved. Machine translations are shown to users in various instances on Facebook and Instagram, including public biographical info and specific pieces of content. In the case of Kannada auto-translation, Meta's AI machine translation model provided a machine translation error which caused inaccurate Kannada translations in Facebook-- which the company claims to have been fixed now.
[2]
Karnataka CM Siddaramaiah Criticizes Meta's Translation Error
Karnataka Chief Minister K. Siddaramaiah reprimanded American tech giant Meta over a faulty translation that falsely declared him dead. He took to X (formerly Twitter), remarking that distorted facts within official communications can damage public trust. For context, the Karnataka Chief Minister's Office (CMO) had uploaded a condolence message for veteran actress B. Saroja Devi on July 15 on Facebook. The CMO account had originally written it in Kannada, and Meta wrongly translated it into English to convey that Siddaramaiah himself had passed away. The erroneous translation on Facebook read: 'Chief Minister Siddaramaiah passed away yesterday, multilingual star, senior actress B. Took darshan of Sarojadevi's earthly body and paid his last respects.' Reacting to the mistranslation, the Karnataka CM tore into Meta's translation capabilities, calling wrong translations "misleading" and "dangerous." "Faulty auto-translation of Kannada content on @Meta platforms is distorting facts & misleading users. This is especially dangerous when it comes to official communications," Siddaramaiah wrote. "Social media platforms must act responsibly. I caution citizens to be aware that translations shown are often inaccurate. Such negligence by tech giants can harm public understanding & trust," he added. Siddaramaiah also informed users that his media advisor had written an official email to Mark Zuckerberg's company over the translation error, asking for its immediate correction. For context, Siddaramaiah's media advisor K.V. Prabhakar asked Meta to temporarily suspend the auto-translation feature for Kannada content. Secondly, he urged the tech giant to consult qualified Kannada language experts and linguists to improve the quality and contextual accuracy of translations between the Dravidian language and English. Meta subsequently fixed the faulty translation, in addition to apologising to the Karnataka CM, as per a Press Trust of India (PTI) report. "We fixed an issue that briefly caused this inaccurate Kannada translation. We apologise that this happened," a Meta spokesperson remarked. The corrected translation now reads: "Chief Minister Siddaramaiah paid his last respects to multilingual star and veteran actress B. Sarojadevi, who passed away yesterday." MediaNama Founder Nikhil Pahwa believes that this mistranslation faux pas by Meta brings the safe harbour debate into focus, and that politicians should figure out resolution mechanisms for such instances rather than lambasting tech companies. He said: "This reaction from Siddaramaiah is indicative of the gap between expectation and reality in terms of how the things work on the internet, and also brings to light the issue of safe harbour for AI. Meta doesn't have people sitting down and translating every piece of text: this is automated using AI. The challenge of translation is that one not only needs to have a cultural context of what is being said, but also a syntactical understanding of languages. As any human translator will tell you, translation isn't done word for word, but has to take into context the overall text, and this isn't easy for machines to do, even though they're getting better. "To test this out, try using Google Translate to first translate from English to another language, and then translate output text back into English, and you'll see that there are errors. Automated translation is done on a best-efforts basis, happens at internet scale, is probabilistic like all AI, and will get better but won't be 100% accurate. How we learn to live in a world with probabilistic outputs by AI, with varying degrees of accuracy, is a problem that government officials and politicians need to first understand and then create mechanisms to resolve, instead of reacting like this, expecting perfection when it isn't possible. The safe harbour need for platforms is only going to increase with AI."
Share
Copy Link
Meta faces criticism from Karnataka's Chief Minister for a faulty auto-translation of Kannada content on Facebook and Instagram, leading to distorted facts and public misleading.
Meta, the parent company of Facebook and Instagram, found itself in hot water after a significant translation error involving Kannada, a language spoken in the Indian state of Karnataka. The incident sparked a controversy when an auto-translated post falsely declared Karnataka's Chief Minister Siddaramaiah dead, leading to a swift response from the state government and an apology from the tech giant 1.
Source: MediaNama
The error occurred when the Karnataka Chief Minister's Office (CMO) posted a condolence message in Kannada for veteran actress B. Saroja Devi on Facebook. Meta's AI-powered translation tool incorrectly translated the Kannada text into English, stating that "Chief Minister Siddaramaiah passed away yesterday" instead of accurately conveying the message about the actress 2.
Chief Minister Siddaramaiah took to X (formerly Twitter) to express his concerns about the faulty auto-translation. He emphasized the dangers of distorting facts in official communications and how it could mislead users and harm public trust 1. The CM's media advisor, K.V. Prabhakar, formally wrote to Meta, urging immediate correction of the issue.
In response to the controversy, a Meta spokesperson stated, "We fixed an issue that briefly caused this inaccurate Kannada translation. We apologize that this happened" 1. The company claims to have resolved the problem, with the corrected translation now accurately reflecting the original message.
The incident has brought attention to the challenges faced by AI-powered translation tools and their potential impact on public communication. Nikhil Pahwa, founder of MediaNama, highlighted that this mistranslation faux pas brings the safe harbor debate into focus 2. He emphasized that automated translation is done on a best-efforts basis and operates at internet scale, making 100% accuracy challenging to achieve.
Source: Economic Times
The Karnataka government has requested Meta to temporarily suspend the auto-translation feature for Kannada content until translation accuracy can be reliably improved 1. Additionally, they have urged the tech giant to consult qualified Kannada language experts and linguists to enhance the quality and contextual accuracy of translations between Kannada and English 2.
This incident underscores the need for tech companies to act responsibly, especially when dealing with official communications. Chief Minister Siddaramaiah cautioned citizens to be aware that translations shown on social media platforms are often inaccurate and that such negligence by tech giants can harm public understanding and trust 1.
As AI-powered translation becomes increasingly prevalent, the incident serves as a reminder of the importance of cultural context and linguistic nuances in automated language processing. It also highlights the ongoing challenges faced by tech companies in balancing the scale of their operations with the need for accuracy and reliability in their services.
Summarized by
Navi
NASA and IBM have developed Surya, an open-source AI model that can predict solar flares and space weather, potentially improving the protection of Earth's critical infrastructure from solar storms.
5 Sources
Technology
7 hrs ago
5 Sources
Technology
7 hrs ago
Meta introduces an AI-driven voice translation feature for Facebook and Instagram creators, enabling automatic dubbing of content from English to Spanish and vice versa, with plans for future language expansions.
8 Sources
Technology
23 hrs ago
8 Sources
Technology
23 hrs ago
OpenAI CEO Sam Altman reveals plans for GPT-6, focusing on memory capabilities to create more personalized and adaptive AI interactions. The upcoming model aims to remember user preferences and conversations, potentially transforming the relationship between humans and AI.
2 Sources
Technology
23 hrs ago
2 Sources
Technology
23 hrs ago
Chinese AI companies DeepSeek and Baidu are making waves in the global AI landscape with their open-source models, challenging the dominance of Western tech giants and potentially reshaping the AI industry.
2 Sources
Technology
7 hrs ago
2 Sources
Technology
7 hrs ago
A comprehensive look at the emerging phenomenon of 'AI psychosis', its impact on mental health, and the growing concerns among experts and tech leaders about the psychological risks associated with AI chatbots.
3 Sources
Technology
7 hrs ago
3 Sources
Technology
7 hrs ago