Bottom line: Remember when the Internet used to be a place for people to discover and share knowledge with others? Those days are fading fast according to cyber security software and services company Imperva.

In its latest annual report, the firm found that a staggering 47.4 percent of all Internet traffic in 2022 came from bots. That is up 5.1 percent from the 42.3 percent of bad bot traffic observed a year earlier. What a waste of resources, no?

Imperva broke down bot traffic into two categories: bad bot traffic and good bot traffic. Simply put, bad bots are those that run automated tasks with malicious intent. They can be used to scrape data from websites without permission or create distributed denial of service (DDoS) attacks. Some bad bots can even facilitate fraud and theft.

Bad bots can be categorized further. In 2022, most – 51.2 percent – of bad bots were deemed advanced in terms of sophistication level compared to simple (33.4 percent) and moderate (15.4 percent) bots.

Bad bots can affect all industries but some are impacted more than others. According to Imperva, gaming sites, telecoms & ISPs, law & government sites, retail pages, financial services, community & society sites, and travel sites were among the hardest hit. Data additionally indicates that bad bots most often target sites in the US. The second largest targeted country, Australia, registered fewer than half as many attacks as the US.

Good bots, on the other hand, are described as those that perform useful tasks like indexing websites for search engines or monitoring website performance.

As bot traffic has increased, human Internet traffic continues to trend downward. According to Imperva, human traffic decreased from 57.7 percent in 2021 to 52.6 percent last year. The high point for real traffic came in 2019 at 62.8 percent but has been going down ever since.

The latest report is Imperva's 10th annual. A lot has changed over the last decade but as illustrated in the chart above, the trend line was not as predictable as one might think.