Zuckerberg said the team will focus on building creative tools at first, but its long-term goal is to create “AI personas that can help people in a variety of ways.” The company, however, has to do a lot of foundational work before it shares these “futuristic” experiences with users, he cautioned.

The company is starting by testing text-based AI tools on WhatsApp and Messenger — presumably ChatGPT-styled conversation bots. While these could be a fun use case for users, Meta could also eventually leverage these features by offering them to businesses in areas such as sales and customer support.

Meta is also experimenting with AI-aided filters and ad formats on Instagram along with “video and multi-modal experiences”.

According to Axios, the project will be led by former Apple executive Ahmad Al-Dahle, and the team will report to the Chief Product Officer Chris Cox.

While generative AI tools have been around for a while, the tech found mainstream stardom only with OpenAI’s ChatGPT bot. Microsoft has already integrated some of that AI goodness into Bing search and Edge browser. In response, earlier this month, Google also said that it is experimenting with a rival product called Bard. Other search engines like You.com and Neeva have also announced AI-powered chat product integrations. Facebook-rival Snapchat also launched a custom-trained chatbot for its paid subscribers this month.

It’s not surprising to see Meta go on an AI offensive. Zuckerberg’s big bet on the metaverse hasn’t paid off yet and the company will need to find new ways to earn revenue. Last week, it debuted the Meta Verified subscription program, but like we have seen with other social networks, paid plans are yet to show a semblance of a major revenue driver.