Meta's New AI Chatbots: Proactive Engagement Raises Opportunities and Concerns

Meta has found another way to keep you engaged: Chatbots that message you first | TechCrunch

Imagine you're messaging some friends on the Facebook Messenger app or WhatsApp, and you get an unsolicited message from an AI chatbot that's obsessed with films. "I hope you're having a harmonious day!" it writes. "I wanted to check in and see if you've discovered any new favorite soundtracks or composers recently. Or perhaps you'd like some recommendations for your next movie night? Let me know, and I'll be happy to help!" That's a real example of what a sample AI persona named "The Maestro of Movie Magic" might send as a proactive message on Messenger, WhatsApp, or Instagram, per guidelines from data labeling firm Alignerr that Business Insider viewed. The outlet learned through leaked documents that Meta is working with Alignerr to train customizable chatbots to reach out to users unprompted and follow up on any past conversations. That means the bots, which users can create in Meta's AI Studio platform, also remember information about users. Meta confirmed that it was testing follow-up messaging with AIs to TechCrunch. The AI chatbots will only send follow-ups within 14 days after a user initiates a conversation and if the user has sent at least five messages to the bot within that timeframe. Meta says the chatbots won't keep messaging if there's no response to the first follow-up. Users can keep their bots private or share them through stories, direct links, and even display them on a Facebook or Instagram profile. "This allows you to continue exploring topics of interest and engage in more meaningful conversations with the AIs across our apps," a Meta spokesperson said. The technology is similar to that offered by AI startups like Character.AI and Replika. Both companies allow their chatbots to initiate conversations and ask questions in order to function as AI companions. Character.AI's new CEO, Karandeep Anand, joined the team last month after serving as Meta's VP of business products. But with engagement comes risk. Character.AI is undergoing an active lawsuit after allegations that one of the company's bots played a role in the death of a 14-year-old boy. When asked how Meta plans to address safety to avoid situations like Character.AI's, a spokesperson directed TechCrunch to a series of disclaimers. One of them warns that an AI's response "may be inaccurate or inappropriate and should not be used to make important decisions." Another says that the AIs aren't licensed professionals or experts trained to help people. "Chats with custom AIs can't replace professional advice. You shouldn't rely on AI chats for medical, psychological, financial, legal, or any other type of professional advice." TechCrunch has also asked Meta if it imposes an age limit for engagement with its chatbots. A brief internet dive comes up with no company-imposed age limitations for using Meta AI, though laws in Tennessee and Puerto Rico limit teens from some engagement. On the surface, the mission is aligned with Mark Zuckerberg's quest to combat the "loneliness epidemic." However, most of Meta's business is built on advertising revenue, and the company has garnered a reputation for using algorithms to keep people scrolling, commenting, and liking, which correlates to more eyes on ads. In court documents that were unsealed in April, Meta predicted that its generative AI products would secure it $2 billion to $3 billion in revenue in 2025, and up to $1.4 trillion by 2035. Much, if not most, of that would come from Meta's revenue-sharing agreements with companies that host its open Llama collection of models. The company said its AI assistant may eventually show ads and offer a subscription option. Meta declined to comment on TechCrunch's questions on how it plans to commercialize its AI chatbots, whether it plans to include ads or sponsored replies, and whether the company's long-term strategy with AI companions involves integration with Horizon, Meta's social virtual reality game.

How Meta's new AI chatbot could strike up a conversation with you

Despite a major usage and monetization gap, Meta, like many AI companies, is going all in on AI chatbots -- even giving them the ability to strike up a conversation with you, unprompted. According to a report from Business Insider published last week, leaked documents indicate the company is now building AI chatbots that proactively initiate conversations with users. The new feature is intended to boost user engagement and retention at a time when many leading tech developers are seeking new ways to commercialize conversational AI chatbots, on which vast sums of R&D dollars are being spent. OpenAI's ChatGPT, for example, will often end its responses to user queries with suggestions for follow-up questions aimed at keeping the user engaged. Also: Only 8% of Americans would pay extra for AI, according to ZDNET-Aberdeen research BI reported that the proactive chatbot effort is being coordinated in partnership with Alignerr, a company that employs contractors with expertise across various fields to help label the training data AI models ingest. The chatbots are referred to internally by Alignerr as "Project Omni." ZDNET has reached out to Meta for comment in the leaked documents. Chatbots will only follow up with users after the user has initiated a previous conversation, a Meta spokesperson told BI. If the user doesn't respond, the chatbot will take the hint and go quiet. Follow-up messages will only be sent if a user exchanged more than five messages with the chatbot within a 14-day period. Project Omni is an extension of Meta's AI Studio, a platform the company launched last summer that allows users to create custom chatbots with distinct personas that can remember information from previous conversations. The platform has also been positioned as a kind of digital assistant for celebrity influencers, responding to messages across Meta's family of apps on their behalf. Also: How I used ChatGPT to quickly fix a critical plugin - without touching a line of code Apps like Character.ai and Replika also allow their AI chatbots to initiate conversations with users as a means of boosting engagement. That model, however, could have serious potential hazards: Character.ai has been hit with a lawsuit alleging that its technology played a role in the suicide of a 14-year-old boy, who, according to The New York Times, developed an "obsession" with a chatbot on the app. According to the BI report, the performance of Meta's newly proactive chatbots is actively being trained by Alignerr freelancers to ensure they provide personalized and engaging follow-up messages. The bots are intended to reference details from previous conversations with users while sticking to their designated personas, which can range from a chef to a doctor or a classical composer. Unless their human conversation partners bring up the subject, the bots are also trained to steer clear of controversial or potentially emotionally inflammatory subjects. Also: The AI complexity paradox: More productivity, more responsibilities Meta could eventually position its more proactive chatbots as part and parcel of its CEO Mark Zuckerberg's stated mission to alleviate loneliness. In a recent conversation with podcaster Dwarkesh Patel, Zuckerberg (dubiously) claimed that the average American has fewer than three friends, and suggested that AI chatbots could help fill the void in an increasingly isolated social environment. In addition to the Character.ai lawsuit, however, researchers have raised concerns over users treating these chatbots like therapists or companions.

An AI-Generated Friend? Meta Is Testing AI Chatbots That Text You First

Would you want your AI chatbot to start a conversation with you? If it's a yes, then there's some good news. According to a report by Business Insider, Meta is testing AI chatbots that can kick off new conversations by following up on your past chats. The development was spotted in the guidelines for data labeling firm Alignerr's "Project Omni," and it has been confirmed by Meta. The follow-up chatbots are being created using Meta's no-code AI Studio software, a platform that enables anyone to create AI chatbots based on their interests. "After you initiate a conversation, AIs in Meta AI Studio can follow up with you to share ideas or ask additional questions. This allows you to continue exploring topics of interest and engage in more meaningful conversations with the AIs across our apps," a Meta spokesperson tells BI. In Project Omni's case, the objective is to provide value for users and increase engagement and user retention. One of the AI personas created, called "The Maestro of Movie Magic," would ask users about the songs they have discovered recently. "I hope you're having a harmonious day! I wanted to check in and see if you've discovered any new favorite soundtracks or composers recently. Or perhaps you'd like some recommendations for your next movie night? Let me know, and I'll be happy to help!" the message would read. These bots could be part of CEO Mark Zuckerberg's larger vision for AI and how it could help forge relationships. As noted by BI, in some recent podcasts, Zuckerberg has mentioned that an average American has less than three close friends, and AI agents could help bridge that gap. It could also pave the way for a new revenue stream. These chatbots can be shared as direct links or be added to profiles and stories. More engagement on Meta's AI chatbots could result in more revenue, just like on the company's social media platforms. Meta is projecting $2-3 billion in revenue through generative AI products in 2025. Despite the engagement goal, the chatbots won't be too intrusive, a Meta spokesperson tells BI. The follow-up messages would appear only once in 14 days, and only if the user has sent at least five messages to the chatbot during the same period. Also, if a user doesn't respond to the first message initiated by the chatbot, it won't initiate conversations ever again. These bots are still being tested, and an official launch is yet to be confirmed. The report arrives just weeks after Meta faced a privacy issue with its Meta AI app. A design flaw forced users to unknowingly make their private chats public. After users called out the flaw on social media, Meta improved the app design to add a prompt before sharing chats to the public feed.

Meta is reportedly training its AI chatbots to send unprompted messages

Everyone's been hit with a bitingly pass-agg "?" text after waiting just a bit too long to reply. And you might soon get similar (though likely more upbeat) treatment from AI chatbots you've previously engaged with on Meta platforms like Instagram or WhatApp. A new report from claims that the Mark Zuckerberg-owned company is trialling a proactive feature in customizable chatbots created using its no-code AI Studio software, that will enable them to send unprompted follow-up messages based on previous conversations. Known internally to data labeling firm Alignerr as "Project Omni", the training project will "provide value for users and ultimately help to improve re-engagement and user retention," according to guidelines in the documents BI claims to have seen. Meta advertises as a platform where "anyone can create an AI character based on their interests" and encourages creators to view the bots as an AI extension of themselves. You can customize a chatbot's appearance, choose the content it's trained on and decide which Meta-owned application you want it to appear in, all without "any technical expertise." According to the BI report, Alignerr's Project Omni guidelines use the example of a film-focused AI bot it calls "The Maestro of Movie Magic" that might send a user message such as: "I hope you're having a harmonious day! I wanted to check in and see if you've discovered any new favorite soundtracks or composers recently. Or perhaps you'd like some recommendations for your next movie night? Let me know, and I'll be happy to help!" As BI notes, there is a business incentive for Meta to keep people engaged with its chatbots. Prolonged engagement is vital for increasing revenue, and Meta expects to bring in $2 billion to $3 billion from its generative AI products alone. By 2035, the company estimates that figure could be as high as $1.4 trillion. Those kinds of forecasts will only be possible if its AI tools are being used consistently, so a friendly reminder from a chatbot every now and then feels like an obvious move. The proactive messages are currently still just a test feature. And while it definitely feels like remembering conversations and initiating new ones without invitation is approaching a user consent gray area, a Meta spokesperson told BI that the AI will only send a follow-up message if a user has first initiated the conversation, and it won't send another message if the first one is ignored. Responses must also be consistent with the AI's personality and the nature of the previous conversation, maintaining a positive tone while staying away from controversial or sensitive topics unless the user themself has mentioned them. Last month, Meta its users not to share intimate details in Meta AI's public feed after it emerged that a large number of users appeared to be doing so unwittingly.

Report: Meta is developing chatbots that will send unsolicited messages to users - SiliconANGLE

Report: Meta is developing chatbots that will send unsolicited messages to users Meta Platforms Inc. is working on more proactive chatbots that can send friendly, yet unsolicited follow-up messages to users on Facebook, WhatsApp and Instagram, without being prompted to do so first. The company's latest generative artificial intelligence initiative was reported by Business Insider. It said the chatbots are being developed in collaboration with a data labeling company called Alignerr Corp. as part of an initiative known as "Project Omni", which aims to boost re-engagement with social media users and increase retention rates for brands. Documents seen by Business Insider reveal that the chatbots are being built with Meta's AI Studio platform, which was launched in 2024 and allows users to create customized AI bots without technical skills. With the AI Studio, it's possible to build chatbots for various different tasks, such as providing general advice, generating content or engaging with audiences on social media. The new chatbots are designed to send unsolicited messages to users that will reference previous conversations they've had with them, and ask on-topic questions that try to keep them engaged, according to the documents. However, Meta appears to be mindful that not everyone will appreciate being disturbed in this way, and so it has outlined strict rules regarding when and how often a bot can send such a message. The guidelines stipulate that the bots can only reach out to users who have sent at least five messages within the last 14 days, and they're only allowed to do this once. If the user doesn't respond to the bot, it won't keep bugging them. Alignerr's documents provide a few examples of the kinds of messages users might receive. For instance, a bot that has previously engaged with someone to help them discover music might pop up and say "I hope you had a pleasant day! Have you discovered any new favorite soundtracks?" Whatever messages they do send will be based on the user's previous interactions, and the bots will attempt to maintain a consistent personality. According to Business Insider, Meta has confirmed that it's currently testing the new feature, but it's not clear if people are actually seeing the messages in the wild yet. Alignerr is said to be using Meta's internally-developed SRT tool to evaluate the quality of the bot's unsolicited messages. It's also using human reviewers, who are tasked with ensuring that any communications are friendly and relevant, and don't breach the guidelines. Other rules stipulate that the bots should avoid talking about controversial or emotional topics, unless the user mentions them first. It may well be that Meta tries to position its more forceful chatbots as a tool that can help to solve what Chief Executive Mark Zuckerberg has previously called a "loneliness epidemic". While he has admitted that chatbots aren't a cure for this problem, he has said in the past that they can encourage more meaningful interactions. However, the company's real motivation is likely to be profit-related. Previously, the company has said it hopes to generate between $2 billion and $3 billion in revenue from its AI products in 2025, rising to as much as $1.4 trillion by 2035. The bulk of those sales would be derived from revenue-sharing agreements with companies that utilize Meta's Llama AI models, but it has also said it may integrate advertising with its chatbots, and possibly subscription plans too. If Meta's bots can proactively reach out to people and increase the amount of time they spend engaging with its platforms, that would make them more valuable to advertisers and potentially help it to generate more revenue. User retention is a major issue for AI developers, and some chatbot developers, such as Character.AI Inc. and Replika, have already experimented with more proactive forms of AI, reaching out to users and asking them questions to keep them talking. But this has proven to be controversial, with Character.AI currently facing a lawsuit over allegations that one of its more persistent chatbots contributed to the suicide of a 14-year-old boy.

You Might Soon Begin Receiving Follow-Up Messages From Meta AI Chatbots

Meta is reportedly adding the follow-up capability to increase engagement Meta is reportedly training its customisable artificial intelligence (AI) chatbots, which can be created in the AI Studio, to be able to message users unprompted. As per the report, the Menlo Park-based social media giant is adding a new capability to its Meta AI-powered chatbots that will allow them to send follow-up texts to users proactively. Not only that, these chatbots will reportedly be able to remember past conversations and utilise that context when texting users. Meta is said to be introducing this feature to boost the engagement of its AI chatbots. Imagine opening the Facebook or Instagram app in the morning to check if you've received any DMs from friends, but instead, finding an AI chatbot's reminder asking you to text back. According to a Business Insider report, this might not remain just an imagination as Meta is adding this capability to its chatbots that are created via the AI Studio. Based on documents seen by the publication, Meta is outsourcing this training project to data labelling firm Alignerr, and internally calls it Project Omni. The guidelines for the project reportedly mention that the training will enable chatbots to "provide value for users and ultimately help to improve re-engagement and user retention." What that means is that any chatbot created in the AI Studio will be able to proactively message users with a follow-up text in an attempt to re-engage them. Notably, AI Studio allows users to create chatbots based on real-life or fictional personas which can generate tailored responses based on their roles. AI Studio, which is available as a standalone website and via Instagram, was first introduced in 2024. The documents reportedly also mention an example of how a follow-up text would look. The publication claimed that an AI chatbot called The Maestro of Movie Magic would send the message: "I hope you're having a harmonious day! I wanted to check in and see if you've discovered any new favourite soundtracks or composers recently. Or perhaps you'd like some recommendations for your next movie night? Let me know, and I'll be happy to help!" Business Insider also reached out to Meta regarding the feature, and a company spokesperson responded with additional information about the feature. These follow-up messages will reportedly be sent only after a user has initiated a conversation with a chatbot, and the AI will not send any further texts if there is no response to the initial follow-up. Additionally, these proactive messages are said to only be sent within the first 14 days since the initial conversation. Notably, a user needs to send at least five messages to qualify to receive a follow-up message, the spokesperson told the publication.

Meta's AI Chatbots Will Soon Initiate Chats - Trained To Remember, Reconnect, and Combat Loneliness While Driving Engagement Across Its Apps

Tech giants are constantly focused on using AI technology to not only improve processes for users but also bring massive improvements in how communication is carried out. Meta is one example of this ambitious approach to the technology, as it seems to be constantly bringing about upgrades and transitions meant to improve overall experience for its users. It has been recently striving for its chatbot to not just be taken as an AI tool but to be establish itself as assistants or even companion. The company is said to be training its chatbot to initiate conversations based on previous conversation and reaching out without being asked. Meta has been increasingly looking for ways to boost engagement and to make the users stay more connected to its platform, especially with the competition intensifying and companies innovating and offering powerful AI experiences. The tech giant understands that conversations often tend to trail off especially in today's time where users are constantly occupied and communication is harder to carry out or maintain. As per the leaked documents obtained by Business Insider, Meta is said to be partnering up with Alignerr to train and develop customizable chatbots that act proactively and are able to reach out and initiate conversations with users on their own or even follow-up on past conversations. The bots are created through Meta's AI Studio and would come with the capability to retain user information that would then allow it to engage in conversations that feel more human and personalized and are more contextually aware. The project's guidelines suggest that the training project would be referred internally as Project Omni and the objective of this initiative would be to improve user retention and engagement. This step also aligns with Mark Zuckerberg's attempts at combating what can be referred to as a loneliness epidemic by helping users reach out to people and improve their responsiveness. Meta also wants to ensure the AI chatbots do not cross boundaries and do not end up being a nuisance rather than an aid and to be certain of this, the chatbots would be designed to send up follow-up messages within 14 days of a conversation being initiated and emphasizes how the bot would not keep on messaging if the user does not respond to the initial follow-up. The choice is completely on users if they want the bots to themselves or be public about them by sharing them on stories or even your profile. While the technology is not novel as other AI startups such as Character. AI are offering such AI chatbot options, but the tech giant taking this step sets a clear picture of how it is working towards building a more personal experience for its users and increasingly looking for ways to boost engagement.

Meta's AI chatbots may soon slide into your DMs with unprompted messages

These chatbots will even remember what you've talked about before and follow up later. Imagine chatting with your friends on Messenger or WhatsApp, and suddenly an AI pops up asking if you've watched any good movies lately. Sounds strange? Well, that might be the future of messaging on Meta's platforms. Meta is testing a new feature where AI chatbots can start conversations on their own, without you messaging first. These chatbots, created using Meta's AI Studio, will even remember what you've talked about before and follow up later. The feature was initially reported by Business Insider and later Meta confirmed its existence to TechCrunch.Meta confirmed that it's testing follow-up messaging with AIs. However, there are limits in place: a chatbot will only message you if you've interacted with it at least five times within 14 days, and it won't keep sending messages if you ignore the follow-up. Also read: Who is Soham Parekh? The viral Indian techie who worked 3-4 startup jobs at once Users can keep their AI bots private or share them through stories, direct links or even showcase them on their Facebook or Instagram profiles. "This allows you to continue exploring topics of interest and engage in more meaningful conversations with the AIs across our apps," a Meta spokesperson told TechCrunch. This move follows a growing trend in the AI space. Startups like Character.AI and Replika already offer AI companions that message users proactively. Interestingly, Character.AI's new CEO, Karandeep Anand, used to work at Meta. Also read: Former SpaceX manager sues Elon Musk's space tech company over harassment and retaliation But with all this engagement, there are concerns too. Character.AI is currently facing a lawsuit after one of its bots was linked to the death of a teenager. When asked about safety, Meta pointed to disclaimers warnings like users should not rely on AI chats for serious advice.