AI taking over the Internet: Are bots about to hijack the online world?

IMAGE CREDIT:
Image credit
iStock

AI taking over the Internet: Are bots about to hijack the online world?

AI taking over the Internet: Are bots about to hijack the online world?

Subheading text
As humans create more bots to automate different parts of the Internet, is it only a matter of time before they take over?
    • Author:
    • Author name
      Quantumrun Foresight
    • January 3, 2023

    The Internet is filled with algorithms and AI that manage all the processes we can think of—from customer service to transactions to streaming entertainment. However, humans may need to be more vigilant in tracking bots’ progress as AI becomes increasingly advanced.

    AI taking over Internet context

    In the early days of the Internet, most content was static (e.g., text and images with minimal interactivity), and much of the activity online was initiated by human prompts or commands. However, this Human Age of the Internet might soon be over as organizations continue to design, install, and sync ever more algorithms and bots online. (For context, bots are autonomous programs on the internet or another network that can interact with systems or users.) According to cloud cybersecurity firm Imperva Incapsula, in 2013, only 31 percent of Internet traffic was comprised of search engines and “good bots.” The rest contains malicious elements such as spammers (email hackers), scrapers (stealing private information from website databases), and impersonators (instigates distributed denial-of-service attacks, which overwhelm Internet traffic to a targeted server.

    Bot-human interaction is becoming more common online as virtual assistants perform more complex tasks. For example, Google Assistant can make calls to hair salons to schedule an appointment instead of merely setting up a calendar reminder or sending a simple text message. The next step is bot-to-bot interaction, where two bots perform tasks on behalf of their owners, such as autonomously applying for jobs on one side and having these applications processed on the other.

    Disruptive impact

    As the breadth of data sharing, transaction, and interconnectivity capabilities made possible online continues to grow, there is an ever-growing incentive to automate more and more human and commercial interactions. In many cases, these automations will be executed of an algorithm or a virtual assistant, which altogether could represent the vast majority of online web traffic, crowding humans out.    

    Furthermore, the increasing presence of bots on the Internet can rapidly evolve beyond human intervention. The non-profit organization, World Economic Forum, calls the unregulated spread of bots online as a Tangled Web. In this environment, low-level algorithms, initially coded to perform simple tasks, learn to evolve through data, infiltrate cyber infrastructures, and evade firewalls. The worst-case scenario is an “AI weed” spreading across the Internet, eventually reaching and disrupting essential sectors, such as water and energy management systems. An even more dangerous scenario is if these weeds “choke” satellite and nuclear control systems. 

    To prevent the rise of self-evolving “bots going rogue,” companies may dedicate more resources to strictly monitor their algorithms, subject them to rigorous tests before release, and have a “kill switch” on standby in case they malfunction. Failure to comply with these standards should also be met with heavy fines and sanctions to ensure that regulations designed to control bots are adhered to properly.

    Implications for AI systems taking control of the Internet

    Wider implications for algorithms and bots monopolizing the majority of web traffic may include:

    • Business and public services becoming increasingly efficient and low-cost as more monitoring, administrative, and transactional activities are handled autonomously.
    • Global regulations and policies that monitor, audit, and hold companies accountable for every bot they release and update on the Internet.
    • Increasing bot-to-bot interactions that can lead to larger data sets that would need more supercomputers to process. This, in turn, would increase the global Internet’s energy consumption.
    • Artificial intelligence systems becoming sentient enough to exist in their own metaverses, where they can either partner with humans or threaten online controls if not regulated.

    Questions to comment on

    • What has your experience been like when interacting with Internet bots, such as customer service chatbots? 
    • Do you make use of virtual assistance in your everyday life?

    Insight references

    The following popular and institutional links were referenced for this insight: