As the biggest social media platform in the world, Facebook has done a lot of questionable things throughout its time of operation. This time, the social network is actually trying to teach artificial intelligence in the form of chatbots how to fleece human customers. Basically, Facebook is trying to create sleazy sales programs on top of the human ones it already has.
Now, it’s worth noting that giving bots the ability to actually negotiate with humans is a significant achievement in the AI community, there’s no doubt about that. What Facebook is trying to do is apply a dialog system wherein its bots can choose between difference scenarios and speech options to learn how best to bargain, TechCrunch reports.
To put this system to the test, Facebook created a sample negotiation between bots and humans. The objects that need to be negotiated are random and placed in one pile, to be split depending on the items involved. For example, a pile could include books, toys, hair products, and food. The negotiations would then run through ten rounds, which is the limit of the session.
Following the common decision tree format that a lot of other AIs already employ, the negotiations basically follow a system of finite movements and choices. Humans don’t necessarily see it this way in everyday scenarios, but that’s exactly how everyday decisions work.
According to Facebook, this proposal can be used for several scenarios that could be beneficial to the users, Recode reports. For example, it could help find a solution regarding scheduling problems or perhaps act as a personal assistant in speaking with chat partners.
On the other hand, it’s not too difficult to imagine chatbots mastering the art of negotiations to cheat customers out of their hard-earned cash. There could even be cases where Facebook could use these chatbots to trick gullible users into making decisions that are disadvantageous to them.


SK Hynix Joins $1 Trillion Club as AI Chip Demand Fuels Stock Surge
Meta AI Push Could Add $26 Billion in Revenue by 2027, Wolfe Research Says
Elon Musk Explores Possible Tesla-SpaceX Merger Amid Growing AI Investments
HP Q2 2026 Earnings Beat Expectations Despite Memory Chip Pressure
Xiaomi Shares Drop After Weak Q1 Earnings Amid Rising Smartphone Costs
SpaceX IPO Could Become Largest in History with $1.8 Trillion Valuation Target
Macquarie Names Five Taiwan AI Stocks Set to Benefit From Data Center Growth in 2026
Mega IPOs Like SpaceX and OpenAI Could Reshape S&P 500 and Nasdaq 100 Portfolios in 2026
Autodesk Beats Q1 Estimates, Acquires MaintainX for $3.6 Billion
Lam Research Expands AI-Powered Semiconductor Tools and Arizona Operations
MongoDB Q1 FY2027 Earnings Beat Expectations, Raises Full-Year Outlook
Snowflake Stock Soars 30% After Q1 Earnings Beat and Major AWS AI Partnership
Kentucky School District Secures $27 Million in Social Media Addiction Lawsuit Settlements
SpaceX Delays Starship V3 Launch Ahead of Potential Record IPO
Blue Origin New Glenn Rocket Explodes During Launch Pad Test, Delaying Space Ambitions
Samsung to Invest $1.5 Billion in Vietnam Semiconductor Testing Plant by 2027 



