OpenAI’s ChatGPT Agent Outsmarts CAPTCHA - 1

Image by Jonathan Kemper from Unsplash

OpenAI’s ChatGPT Agent Outsmarts CAPTCHA

  • Written by Kiara Fabbri Former Tech News Writer
  • Fact-Checked by Sarah Frazier Former Content Manager

OpenAI’s ChatGPT Agent successfully completed the human verification test by simply clicking on it, as the tech community watched with surprise.

In a rush? Here are the quick facts:

  • The AI narrated the verification process in real time.
  • Cloudflare’s system didn’t trigger a full CAPTCHA challenge.
  • Reddit users shared screenshots of the AI completing tasks online.

The ChatGPT Agent from OpenAI passed an anti-bot verification test by clicking on the “Verify you are human” checkbox, which left the tech community surprised, as first reported by ArsTechnica .

In a screenshot shared on Reddit, the AI passed Cloudflare’s “Verify you are human” checkbox and later explained its step-by-step actions.

“The link is inserted, so now I’ll click the ‘Verify you are human’ checkbox to complete the verification on Cloudflare. This step is necessary to prove I’m not a bot and proceed with the action,” the bot stated as it moved through the process, as reported by ArsTechnica

The ChatGPT Agent works as a new OpenAI AI assistant browsing tool that operates in an isolated virtual environment. The system requires user authorization before it can execute actual operations, including buying products. The irony of this situation became apparent as the bot successfully proved its “not a bot” status.

The agent skipped the image-based CAPTCHA but successfully passed the Turnstile system of Cloudflare, which evaluates human interaction, which uses mouse movements, browser behavior, and other cues to test for human interaction.

The ability of bots to bypass these systems isn’t entirely new, but ChatGPT Agent’s live narration adds a surreal touch. CAPTCHA systems, meant to separate bots from humans, are now sometimes helping to train the AI they’re meant to stop.

Some people believe this development will become a standard in the future. The security tests become easy for ChatGPT Agent to pass, yet it faces difficulties with poorly designed websites, proving that bad web design might be more effective than CAPTCHA at stopping the bots.

ChatGPT Conversations Temporarily Indexed On Google - 2

Photo by Solen Feyissa on Unsplash

ChatGPT Conversations Temporarily Indexed On Google

  • Written by Andrea Miliani Former Tech News Expert
  • Fact-Checked by Sarah Frazier Former Content Manager

OpenAI temporarily released a feature that allowed users to make private conversations public and discoverable on Google. However, it was quickly removed after concerns arose that users were unknowingly sharing private conversations.

In a rush? Here are the quick facts:

  • OpenAI removed the feature that allowed users to share conversations publicly.
  • Experts raised concerns over private conversations getting accidentally indexed on Google.
  • The company acknowledged that users could mistakenly share private chats.

Just a few hours after Jarovsky issued her warning, Dane Stuckey, OpenAI’s chief information security officer, announced on X that the feature had been removed.

“We just removed a feature from ChatGPT that allowed users to make their conversations discoverable by search engines, such as Google,” wrote Stuckey. “This was a short-lived experiment to help people discover useful conversations.”

Stuckey explained that users had to opt in to use the feature by selecting a chat to share, checking a box, and choosing to “make it discoverable.” However, the company acknowledged that users could mistakenly share private conversations and decided to disable the feature.

“Ultimately, we think this feature introduced too many opportunities for folks to accidentally share things they didn’t intend to, so we’re removing the option,” added Stuckey. “We’re also working to remove indexed content from the relevant search engines.”

The removal of the feature has already begun to roll out. Stuckey emphasized that user safety remains the company’s top priority.

According to CybersecurityNews , indexed conversations could be found by searching “site:chatgpt.com/share” along with relevant keywords.

This isn’t the first time users have accidentally shared AI interactions with the public. A few months ago, experts raised concerns about users sharing private information on Meta AI’s app through the “discover” option.