Wells Fargo says AI is not in a bubble
We’ve all had the eerie moment of realizing we’re talking to a bot online. Long before ChatGPT made chatbot interactions mainstream, bots were everywhere—from MIT’s ELIZA in 1966 to Microsoft’s infamous Clippy. But by the time of the 2016 election, bots took a darker turn on social media, fueling chaos and misinformation.
Today, bots—automated software that runs repetitive tasks—are still everywhere and play a major role in the AI boom. According to cybersecurity firm Imperva, bots now account for over half of global internet traffic. In 2024, they officially outpaced human-generated traffic. About 20% are classified as "bad bots," responsible for malicious activities like ad fraud and fake engagement.
These bots can skew key web metrics by generating fake clicks, impressions, and user sessions—metrics investors rely on to evaluate companies. Cybersecurity firms estimate the economic impact in the hundreds of billions of dollars annually. Startups often showcase inflated "vanity metrics" like downloads or sign-ups, many of which may be bot-driven and self-reported.
This metric distortion feeds a broader concern: the AI investment bubble. Economist Torsten Slok of Apollo Global Management recently warned that the top 10 companies in the S&P 500 are more overvalued now than they were in the dot-com era. While he didn’t mention bots, the inflated numbers raise the question: is the AI boom riding on artificial traffic?
The rise of “unicorn” startups—private firms valued at over $1 billion—exploded during low-interest years like 2018 and 2021. Now, venture capital has moved heavily into AI, despite questions around what’s real and what’s artificially inflated.
History suggests hype eventually collides with reality. Growing awareness of bot-driven inflation may prompt investor skepticism and regulatory crackdowns. The FTC has already taken steps—banning fake reviews in 2024 and enforcing the BOTS Act, which prevents bots from buying up concert tickets en masse, as seen during Taylor Swift’s Eras Tour. States like California require bots to disclose themselves when trying to influence voters or consumers.
Bots aren’t going away—but as they blur the line between reality and manipulation, the push for accountability is only beginning.