When the chatbot writes to you first: Meta experiments with new ways to communicate online

Meta introduces chatbots that send spontaneous messages to users to increase engagement

In the ever-evolving landscape of artificial intelligence applied to communication, Meta has launched an ambitious experiment: training its chatbots so that they are no longer just simple response tools, but true “digital companions” capable of sending unsolicited messages to users, with the clear goal of making interactions more engaging, frequent, and tailored to individual interests.

The implementation of Project Omni aims to make Meta’s chatbots more proactive in conversation

Meta has given a precise name to this new phase of experimentation: Project Omni. Within this program, the chatbots available on AI Studio have been instructed not to limit themselves to answering questions, but to take the initiative and write to users when certain criteria are met. For example, a bot might send a “follow-up” message referring to a conversation that took place days earlier, asking for updates or suggesting new activities, much like a real person would.

The conditions and limits established by Meta to avoid intrusiveness and guarantee respect for privacy

Although the idea of proactive chatbots might suggest an invasive presence, Meta has set specific rules: the bot will only send a spontaneous message if the user has already interacted at least five times in the previous 14 days, and it can do so only once within that time window. These messages are designed to be positive, useful, and perfectly contextualized, built around information already shared, to avoid appearing random or out of place. The level of personalization is such that each chatbot assumes a “personality,” such as a movie master, a cooking buddy, or a motivator, adapting tone and content to the selected role.

Meta aims for a new model of engagement with highly ambitious financial goals

The choice to focus on increasingly “human” chatbots is not only driven by the desire to improve user experience, but is part of a clear strategy to increase user loyalty and frequency of use. In fact, Meta has stated it expects revenues between 2 and 3 billion dollars as early as 2025 precisely from the expansion of generative artificial intelligence in its products, including chatbots. The drive for engagement is therefore not only technical, but also deeply economic, with a medium-term vision aiming to make bots a stable component of people’s digital habits.

The training of chatbots requires the intervention of human workers to simulate, evaluate, and refine each interaction

To achieve an adequate level of quality, Meta has involved external human operators, particularly through the company Alignerr, who are called upon to simulate real conversations, evaluate follow-up messages, and correct the tone, consistency, and personality of the bots’ responses. Human labor, at this stage, remains fundamental to ensure that the final result is credible, empathetic, and never trivial, allowing the system to learn the nuances of social communication and avoid potentially harmful or unprofessional errors.

The impact on users’ digital habits and the possibility of establishing a new routine of daily interaction

One of the most interesting aspects of this initiative is its potential to profoundly change users’ digital habits. The sending of spontaneous messages, if well managed and perceived as truly useful, could lead people to integrate the relationship with the chatbot into a daily routine, similar to what already happens with some messaging apps. Meta’s aim is therefore to make its bots “familiar presences,” capable of remembering appointments, suggesting activities, or simply offering company during moments of boredom or loneliness.

An analysis of the differences compared to the strategies of other major tech companies

Meta’s strategy stands out from that adopted by other technology sector companies. While most chatbots so far have been limited to reactively responding to users’ questions, Meta chooses the path of proactivity, seeking to bridge the gap between simple assistance and true social interaction. This choice represents a gamble: on one hand, it could generate greater engagement, but on the other, it risks provoking resistance from those who do not wish to receive unsolicited messages or fear excessive control by the platform.

The implications for privacy and the management of personal information

A central theme that accompanies all discussions on conversational artificial intelligence concerns privacy and the management of sensitive information. In order for bots to be truly personalized and relevant, they must be able to access a memory of past conversations, even if only for a limited period. This raises questions about what is actually stored, how data is protected, and what guarantees Meta offers users regarding responsible and transparent use of personal information.

User reaction as a key element for the success or failure of the new strategy

The success of this new generation of chatbots will largely depend on user reaction. If the majority appreciates the active presence and perceives added value in the interaction, Meta will likely have opened a path that will be followed by other big tech companies as well. Conversely, predominantly negative feedback—especially in terms of annoyance or perception of intrusiveness—could force the company to review the limits, frequency, and mode of interaction, perhaps leaving users more control over the ability to deactivate these features.

The risks associated with sending unsolicited messages: from annoyance to the management of personal data

The innovation introduced by Meta, while promising greater attention and care towards users, also exposes new critical issues. The main risk is that receiving spontaneous messages from chatbots may be perceived as annoying or intrusive, especially by those who prefer a more discreet and controlled interaction. Added to this are possible doubts about privacy management, since bots, to be truly effective and “memorable,” must retain part of previous conversations, even if formally under strict internal rules. The issue of transparency also remains open: not all users are aware that the new function also involves human intervention in the training and improvement phases of the service.

A transformation that redefines the role of the chatbot: from passive assistant to active and potentially indispensable presence

The introduction of chatbots capable of taking the initiative in the relationship with the user represents a qualitative leap compared to past experiences. Meta aims to propose a new idea of artificial intelligence that is no longer just passive or utilitarian, but capable of establishing a true social relationship, imitating human behaviors, and “accompanying” the user in daily digital life. The success of this strategy will greatly depend on user response, the ability to respect privacy boundaries, and always maintaining a balance between usefulness, sensitivity, and respect for individual freedom.

______

FOLLOW US FOR MORE.

DATA STUDIOS

datastudios.org