News (EN)

Silicon Valley tech companies replicate online platforms to speed up training of AI agents

Vale do Silício
Vale do Silício - LoveMetaverse/Shutterstock.com

Technology companies in Vale of Silício are intensifying the creation of almost identical replicas of popular websites, such as Amazon and The central objective is to enable AI agents to navigate the internet autonomously, simulating complex interactions in controlled and safe environments. Essa innovative approach, although promising for technological development, has already generated debates and legal notifications for potential copyright violations, raising important discussions in the current legal scenario. The trend points to a significant evolution in the ability to automate online tasks, with projections for 2025 indicating an increase in the presence of these agents on various digital platforms.

Virtual replicas boost autonomy of AI systems

Cloned websites function as controlled environments where artificial intelligence models learn to perform real tasks, such as booking flights, scheduling appointments or making purchases, without interfering with the traffic of the original portals. Desenvolvedores insert variations in scenarios to test different situations that can occur on the real web, optimizing systems resilience.

vale do silício
silicon valley – NorthSky Films/Shutterstock.com
[[_0]

The technique accelerates learning to navigate and execute specific commands, allowing AI agents to develop a deep understanding of digital interfaces. Start-ups state that the method is essential to evolve current chatbots into agents capable of operating independently and efficiently.

Legal actions and the issue of copyright

United Airlines, for example, identified an almost perfect copy of its official website in mid-2024, which maintained buttons, menus and even the company logo. The replica had an identical layout to the official website, including links to frequent flyer programs, which led to quick action from the company.

Legal experts at the airline sent a formal copyright takedown notice to the start-up responsible. Após notice, the technology company changed the domain name to Fly Unified and removed elements of the brand, avoiding a lawsuit.

Prototypes advance to real-use scenarios

AGI, led by Div Garg, maintains a modified version of the United Airlines website for internal testing, focusing on functionality without using the original visual identity. The company removed its original logo and name after receiving notification from the airline, demonstrating its adaptability in the face of legal issues.

Matrices, co-founded by John Qian, develops replicas of several popular online services, ranging from e-commerce platforms to productivity tools. Esta expansion aims to cover a wider range of digital interactions, preparing agents for various eventualities.

Other start-ups follow a similar strategy with e-commerce portals and productivity tools, seeking to refine the capabilities of their agents. The objective is to create systems that can interact with any internet portal in a fluid and effective way, simulating the human experience.

Training in controlled environments minimizes risks

The evolution of AI agents directly depends on training in environments that reproduce the complexity of the real internet, but without the risks inherent in interactions on platforms with active users. Replicas eliminate potential problems such as transaction errors or data manipulation on original websites, which serve millions of people daily.

This simulation method allows developers to exhaustively test the limits and capabilities of AI agents, identifying flaws and improving algorithms in a safe environment. The ability to replicate specific and variable scenarios is crucial for developing robust and reliable systems.

Data security and the integrity of online operations are paramount concerns, and virtual copy training offers a practical solution to mitigate these risks. Replication allows full control over the learning environment, ensuring that errors do not cause real harm.

Companies predict that, with continued advancement, these more sophisticated agents will take on tasks currently performed by administrative-level professionals. The advance represents an important step in the transition from chatbots to autonomous systems with greater capacity and responsibility.

Companies seek to improve interaction and functionality

Developers design systems capable of booking tickets, scheduling meetings, creating reports and managing emails automatically, relieving the workload of human teams. The goal is to create virtual assistants that can operate with a level of autonomy and efficiency that approaches or exceeds human intervention in routine tasks.

Continuous training aims to prepare agents to operate on any internet portal, adapting to different layouts and workflows. The ultimate goal is to create systems that perform complex tasks with minimal human supervision, freeing professionals for more strategic and creative activities.

Legal and ethical challenges in the development of AI

Creating copies of popular websites raises complex legal questions about using layouts, user interfaces, and trademarks without explicit permission. Companhias airlines, e-commerce platforms and banking services are closely monitoring the emergence of unauthorized versions of their portals. The discussion involves the delimitation between use for research and development purposes and infringement of intellectual property, a field still evolving in digital law. Start-ups argue that replicas are private environments, used only for internal development and functionality testing, arguing that there is no intention of competition or fraud. However, visual similarity can create confusion and potentially infringe trademark rights, requiring a delicate balance between innovation and legal compliance.

The rise of freelance agents in the job market

The projection of systems capable of assuming administrative functions indicates a transformation in the labor market.

  • Process optimization:Agentes of AI can speed up repetitive tasks.
  • Cost reduction:Potencial to reduce the need for labor in certain functions.
  • New opportunities:Criação of jobs in AI development and oversight.
  • Professional requalification:Necessidade of adapting the workforce to new technologies.
  • To Top