The use of AI-powered digital assistants during virtual meetings is prohibited by the EU. During a movie conference, these assistants are frequently used to record, take notes, or perhaps report visuals and audio.
A word on the” Online Meeting Etiquette” drop state” No Artificial Agents are allowed,” according to a lecture from the European Commission delivered to European Digital Innovation Hubs earlier this month.
AI agents are tools that you interact with applications like video conferencing software to perform sophisticated, multi-step tasks freely frequently. For instance, Salesforce uses AI to telephone sales prospects.
When questioned by Politico, the Commission stated that this was the first time this law had been implemented, but it did not specify why when questioned why. Although there is no particular EU law that applies to AI officials, the AI models that generate them will have to adhere to the contentious and contentious provisions of the AI Act.
Safety concerns are raised by AI agencies.
According to a 2025 statement from international AI experts, security risks come from the user being aware of what their AI agents are doing, their inherent ability to operate outside of the user’s control, and possible AI-to-AI interactions. However, notetakers and another agent types are not inherently a security threat. Due to these factors, artificial intelligence is less formulaic than conventional designs.
SEE: How Is Artificial Be Used Properly? Scientists From Microsoft, IBM, MIT, and Harvard Alligh In.
Tech firms must be careful when advertising goods that can increase in value without the user’s knowledge. One of the biggest admonitions is about Microsoft Recall, an AI tool that let users control their computers or browse through data in natural language.
Recall saved screenshots of effective windows as a timeline, which raised questions about privacy and data use, and caused significant launch delays. Recall captured screenshots of these at a cost. A number of brokers created by Microsoft to combat virtual threats have since been released.
In more ways than one, AI agencies are becoming more prevalent.
The AI people haven’t stopped giving their designs more control, though. In October of this year, Anthropic made its Claude Sonnet robot available for use with a computer use feature that enabled it to move cursors, push buttons, and type text. Its deep study work, which was announced this year, responds to questions “agentically,” just like Microsoft’s equivalent.
In line with their growing relevance, OpenAI expanded its text-to-speech and speech-to-text resources to agentic models next quarter. OpenAI unveiled Operator, an agentic application that runs in-browser to perform tasks like ordering foods or arranging travels in January 2025.
Notice: The EU invests €1.3 billion to increase AI adoption and further enhance “digital competencies.”
Anthropic and OpenAI are also collaborating to advance agent technology, with the latter incorporating support for Model Context Protocol, an open-source standard for connecting AI software, including officials, to data repositories. Additionally, Anthem has collaborated with Databricks to assist big corporations in creating their own agencies.
By the end of 2024, TechRepublic predicted a rise in AI agent apply this year. In a January blog post, OpenAI CEO Sam Altman echoed this statement, saying” we may see the first AI agencies” visit the workforce” and significantly alter the production of businesses.
According to Gartner, 33 % of enterprise software applications will include agentic AI by 2028, an increase from the lowest percentage in 2024. By that year, agents will handle at least 15 % of daily decisions for online stores and a fifth of all interactions with customers.
