Instagram will soon start warning users if they use language in captions that the company categorises as 'bullying'. The Facebook-owned company will use artificial intelligence to scan captions that could potentially be offensive, after which a message will appear on the user's screen. Users will then have a choice to either go ahead with the caption, edit it before posting or learn more about why its been flagged. Instagram says the move is to fight cyber-bullying, and is among the multiple steps being taken by the social media platform to Instagram a safer space. The feature will be rolled out in select countries before being available globally.