what is perplexity's shady thing?
- User
what is perplexity's shady thing?
- Assistant
Perplexity AI has been accused of several questionable practices based on investigations by WIRED and independent developers. The most significant issues include:
Deceptive Web Scraping Practices
Perplexity has been caught ignoring the widely accepted Robots Exclusion Protocol (robots.txt) to scrape content from websites that explicitly attempted to block its crawler[1]. A WIRED investigation discovered that Perplexity was using at least one unpublicized IP address (44.221.181.252) to access restricted content across multiple websites, including WIRED and other Condé Nast publications[1].
When confronted about this practice, Perplexity removed references to its public IP pool from its documentation[1]. Developer Robb Knight independently verified this behavior, finding that Perplexity had ignored his robots.txt file and evaded his firewall to access content he had explicitly blocked from bots[1].
Content Misrepresentation and Hallucinations
Despite marketing itself as providing "instant, reliable answers to any question with complete sources and citations included," Perplexity has been found to:
- Generate fabricated content, including falsely attributing quotes to real people[1]
- Create inaccurate summaries with minimal attribution to original sources[1]
- In one troubling case, falsely claim that WIRED had reported a specific police officer in California had committed a crime (which WIRED never reported)[1]
Misleading Business Description
Perplexity has been inconsistent about what exactly their service is. Their CEO Aravind Srinivas has described it variously as:
- An "answer engine"
- "Like Wikipedia and ChatGPT had a kid"
- A mere "aggregator of information"[1]
This shifting description appears tactical, especially when responding to accusations about copyright infringement or content misuse.
Potential Copyright Infringement
Publishers including Forbes have accused Perplexity of effectively republishing their original reporting with minimal attribution[1]. Forbes' general counsel sent a letter to Perplexity demanding they remove misleading articles and compensate them for alleged copyright infringement[1].
Harmful Impact on Original Content Creators
By creating comprehensive summaries of original content, Perplexity potentially diverts traffic from the original sources. This threatens the business model of publishers and content creators who rely on site visits for revenue[7]. As one commenter on Hacker News noted, this practice "decreases the probability that users would come to [the original] content" which is "unacceptable" for content creators[7].
The WIRED investigation concluded that "the magic trick that's made Perplexity worth 10 figures, in other words, appears to be that it's both doing what it says it isn't and not doing what it says it is."[1]
Other models, same or similar prompt
Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.