- The Writing Help feature is first being rolled out on WhatsApp in the U.S.
- The feature is set to reach other locations in the coming days, WhatsApp said
- For now, the AI writing is only being made available for messages in English
Hopping onboard the generative AI trend, Meta has started rolling out the “Writing Help” feature on WhatsApp. The feature will be accessible within the messaging app under a new “pencil” icon, WhatsApp said in an official blog post this week.
“You can review the suggestions from AI in various styles such as professional, funny, or supportive that you can select or continue editing to deliver that perfect message,” the platform said, announcing the development.
As WhatsApp claims to maintain an end-to-end encryption, it asserted the same level of privacy for this suggestive AI feature as well.
Newsletter
Get weekly updates on the newest crypto stories, case studies and tips right in your mailbox.
“Writing Help is built on top of Private Processing technology, which allows you to leverage Meta AI to generate a response without Meta or WhatsApp ever reading your message or the suggested re-writes,” it highlighted.
Meta’s ‘Private Processing’ twist for WhatsApp
Essentially, the Private Processing technology would create a secure cloud environment for WhatsApp AI features to analyze and process data without exposing it to unauthorized parties.
Looking for users to have more control over their AI uses, WhatsApp said that Private Processing features like Message Summaries and Writing Help are off by default. Those interested to use them can do so manually from the settings. A blog by Engineering at Meta has also noted that Private Processing is being developed as an optional capability.
Meta engineers are building the Private Processing technology upon a Trusted Execution Environment (TEE), allowing users to access AI features within WhatsApp’s private cloud environment.
Confidential processing, enforceable guarantees, and verifiable transparency have been marked as the three core principles guiding the development of this technology, Meta claims.
To add more layers to secure its ecosystem, Meta said, it is deploying more safety measures to keep threat actors away from the Private Processing technology
“An attacker should not be able to target a particular user for compromise without attempting to compromise the entire Private Processing system. Private Processing must not retain access to user messages once the session is complete to ensure that the attacker can not gain access to historical requests or responses,” the engineers’ blog said, outlining the core security ideas.
Over the last few months, multiple widely used platforms like Google and X among others have added similar AI writing and summary features to their platforms to keep up with the advancing technologies while also making these part of people’s day-to-day web interactions.
