Meta's new 'optional' AI chatbot in apps like Whatsapp is proving ...
You may have noticed in recent weeks the appearance of a benign blue-pink-green-hued circle in all your Meta-owned apps and perhaps wondered what it is. But other users have already expressed annoyance about this new artificial intelligence (AI) feature in WhatsApp, Facebook, and Instagram, especially since it can’t be deleted. The AI assistant has been rolled out on Meta apps since March, but the move hasn’t been smooth sailing.
AI Assistant by Meta
The Meta AI is a conversational agent (or chatbot) powered by AI. More precisely, it is powered by Llama, the large language model (LLM) developed by the company. "Whether you’re planning a group trip, brainstorming dinner ideas, settling an ongoing debate, or just looking to add some fun to your chat, Meta AI is your on-call assistant," the company said. Meta AI works in the same way as any other chatbot does by responding to the queries the users send it.
A lot of users expressed their annoyance at the feature being added by default and hard - if not impossible - to remove, prompting discussions on a Reddit subreddit about how to disable it on WhatsApp. Some users reached out last month to MEP Veronika Cifrová Ostrihoňová from the social-liberal Progressive party in Slovakia, who has since asked the Commission if the feature aligned with EU rules.
Concerns and Controversies
Overall, a rising number of people are expressing a sort of "AI fatigue" amid the constant rollout of new applications or new features. Meta was already under fire for the "opt-in by default" feeding of its model with users’ data based in Europe unless they explicitly refused to have their data used. Sardeli added that Meta wasn’t "transparent enough" regarding how it was using the users’ data. "AI technologies" is an umbrella term that includes a wide range of technologies, which are mostly unknown to the average user of Meta products," the lawyer from NOYB said. "Meta has an obligation to inform its users about exactly what it does with their personal data, an obligation which it is currently trying its best to avoid," she added.
Options for Users
The feature can’t be disabled; however, it’s possible to mute the AI chat in WhatsApp by long pressing on the chat and tapping the option. To opt-out of the data training, users can submit an objection request via the dedicated form. Though some websites advertise downgrading the version on Android - it can’t be done on iOS - by downloading an old version of the app, this solution isn’t recommended for security reasons.
Expansion and Accessibility
Meta AI was rolled out as a new feature across Facebook, Instagram, WhatsApp, and Messenger, the applications owned by Meta. Launched first in the US in September 2023, it became accessible to European users this year. The circle icon is available to users in six languages: French, German, Hindi, Italian, Portuguese, and Spanish. It’s also available on Ray-Ban smart glasses in several European countries, including Germany, Austria, and Belgium.
Conclusion
AI companies like OpenAI and Meta use data made by others to train their AI models. The US Copyright Office says, in some cases, that may be illegal. UNESCO’s approach to regulating neurotechnology considers future applications and current privacy and consent concerns. Many of the world’s languages are at risk of extinction. However, researchers are developing AI tools they hope can safeguard languages for future generations.
