Meta, the parent company of Facebook, Instagram, and WhatsApp, has implemented its artificial intelligence (AI) in the instant messaging app, expanding its functionalities. The technology, which had been available in other countries since April 2024, took a bit longer to arrive in Brazil due to requirements from the National Data Protection Authority (ANPD).
WhatsApp’s AI is based on advanced language models like LLaMA (short for Large Language Model Meta AI), trained on vast amounts of textual data to effectively understand and generate natural language. “Meta’s AI can answer questions, provide recommendations, search the web for news on topics of interest without leaving the app, and generate images and small GIFs for sharing,” explains Pierre dos Santos, AI Analyst at Leste Telecom.
“However, the tool is still in Beta, so it contains many usage errors. This will improve over time, and the AI’s functionality may even be revised as there are many opportunities to add new services, including those related to accessibility,” he adds.
Hero or villain? Depends on usage
With so much debate about artificial intelligence, which has already been proven behind practices like fake news and deepfakes, many people are apprehensive about Meta’s AI being available on WhatsApp without even the option to disable it. “Meta has stated that the content of conversations with the AI may be used to train artificial intelligence algorithms, but it does not associate this content with users’ personal information,” reassures Pierre.
Although it hasn’t publicly disclosed how the data will be used for ad targeting beyond AI training, the continued use of the tool could impact the types of ads users receive in the long term. Data collection, a common practice in the tech industry, may be used for ad personalization, audience segmentation, and behavior prediction, for example.
“However, my expectation is that Meta will prioritize user privacy and consent, using AI ethically and transparently to benefit both users and advertisers in line with our legislation,” the analyst opines.
Even though the technology doesn’t have access to private WhatsApp conversations and user data remains protected by the messenger’s encryption, according to the AI documentation, messages shared with the tool may be used to provide relevant answers for you or to improve the technology. “Therefore, don’t send messages containing information you don’t want to share with the AI. At the very least, we can delete messages sent to the AI by typing /reset-all-ais in the conversation,” he warns.
Use sparingly
Pierre also states that AI is a powerful tool that can be useful in many contexts. However, it’s essential to use it responsibly and carefully, always considering personal data security and privacy. To this end, he shares some basic but valuable tips:
- Use AI as an assistant tool, not as a substitute for critical thinking;
- Use AI for tasks you consider safe and without privacy risks, avoiding sharing personal or confidential information with the AI in conversation;
- Avoid using AI for making important decisions;
- Only research general interest topics, avoiding sensitive or controversial subjects.
“It’s true that it’s becoming increasingly difficult to identify whether content is AI-generated, but there are some telltale signs: unknown or dubious source; too-good-to-be-true content; lack of authorship information; artificial language; generic and unoriginal content; and lack of emotion and subjectivity,” concludes the expert.
