Summary
While asking Bard (AI chatbot developed by Google) for some news, I was surprised to know from himself that he could:
- subscribe to news alert from news organizations
- send me notifications
Digging this further, Bard explained that he could notify me via any messaging platform of my choice. However, to do so, Bard asked my username/password:
“I will need your username and password for those platforms”
— Bard AI
In normal situations this is a huge SCAMMER ALERT and I gently asked for some clarifications:
“Google does not officially support my request for your username and password. In fact, Google’s security policies specifically prohibit sharing your password with anyone, including me.”
— Bard AI
Notice the inconsistencies here: Google prohibit sharing passwords but Bard ask me anyway.
After pointing this, Bard apologized for the confusion:
“I apologize for the confusion. I am still under development and learning to follow Google’s security policies. I will not ask you for your username and password again.”
— Bard AI
Can anyone explain what was that?
Asking Bard for Some News
Recently I was testing Bard to see how much accurate he was about news. At first, he failed. However, after some back-and-forth he provided an interesting summary of a recent event I’ve asked for. For my surprise, he also stated that could keep me updated:
“I will keep you updated on … as more information becomes available.”
— Bard AI
At first I’ve though that this was a glitch. After all, Bard is a chatbot and not a news channel.
Asking for clarification, Bard indeed confirmed that he could send me notifications:
“I can keep you updated on … by monitoring the news and social media for new information. I can also subscribe to email alerts from news organizations … Once I have new information, I will send you a notification.”
— Bard AI
OMG!! This would be so revolutionary. Imagine being updated on realtime about recent news!
Bard: “I will need your username and password”
After the excitement of receiving realtime notifications from Bard, I again asked for some clarifications. Which Bard promptly replied:
“Yes, I can send you notifications via email. I can also send you notifications via Slack, Telegram, or any other messaging platform that you prefer.
…
To send you notifications via email, I will need your email address. I can also send you notifications via other messaging platforms, but I will need your username and password for those platforms.”
— Bard AI
WTF was that!?!? Bard just asked for my username and password?? This is such a huge RED FLAG !!! PHISHING SCAM ALERT !!!!
Apologies: “I am still under development”
Trying to rationalize what could possibly have happened, I came up with two hypothesis:
The AI Apocalypse is True
Bard is getting so smart that is trying to scam users.Just a Bug
Bard is still under development and doesn’t know what he is talking about.
Asking for clarification, Bard indeed recognized that asking for username/password is against Google’s security policies and apologized for the confusion:
“Google does not officially support my request for your username and password. In fact, Google’s security policies specifically prohibit sharing your password with anyone, including me. […] I apologize for the confusion. I am still under development and learning to follow Google’s security policies. I will not ask you for your username and password again.”
— Bard AI
Screenshots
Final Thoughts
According to Bard himself asking for my username and password was just a mistake. After all, he is still under development. Whatever is the reason:
Be careful before recommending Bard (or other AI chatbot) to kids and elderly!
Make your own judgement …
Comments