You can enable subtitles (captions) in the video player
[MUSIC PLAYING]
INTERVIEWER 1: Robots are making even more of a mark on banks trading floors, and not just by wheeling and dealing at the speed of light. Banks and trading houses are increasingly using artificial intelligence programmes to monitor traders. They're doing this so they can spot signs that one of them might be considering a rogue trade that could leave their employer with a big and expensive headache. AI looks at everything from how traders phrase their emails, to the times they use their computers.
If they use advanced analytics techniques to identify potential red flags, the artificial intelligence aspect means programmes constantly evolve. So they get better at spotting real warning signs instead of lighting of thousands of irrelevant emails. That means that banks compliance teams don't get bogged down in vast amounts of data. Behavox, a company that specialises in this kind of AI, says that for one client, it reduce the volume of red flags by around 90%.
Computing giant, IBM, says it's in talks with banks about systems that would go even further. IBM's AI, called Watson, can look at everything from leader convictions to divorce settlements and other publicly available information. That means some with a big divorce settlement could be flagged as a higher risk for rogue trading since they might need the money, as can someone with a court conviction that might be evidence of dishonesty. In the last decade alone, the world's top 13 banks have paid more than $13 billion in fines and losses linked to rogue trading. That means that anyone selling these services to banks is pushing at an open door, provided, of course, they can demonstrate the tools work.
[MUSIC PLAYING]