The world's largest video-sharing website has begun public testing of its YouTube Video Identification Technology, following a private test with nine media companies.
Only two of the media companies, Walt Disney and Time Warner, have given the Google-owned website permission to talk about their involvement.
The new technology is based on a database that stores reference files of original video content and associated ownership rights.
When a YouTube user uploads a video, the database scans the content and cross-references it against the files stored in its database for copyright infringement. The software can scan videos ranging from 10-minute clips to three-hour movies.
YouTube is considering making its database available to other video sites, a move that would eliminate the need for media owners to work with different copyright protection systems on multiple video-sharing websites.
YouTube has been under increasing pressure from traditional media companies, which have accused the website of dragging its heels and failing to do anything to offer reliable ways to identify video clips uploaded by users without permission.
In March, Viacom launched legal action against YouTube, seeking $1bn in damages.
The media owner has charged the website with "massive intentional copyright infringement", after demanding the removal of clips of its popular shows 'The Colbert Report' and 'The Daily Show with John Stewart'.
David King, YouTube product manager, and Glenn Brown, partner development manager, said in a joint posting on the company's website: "We expect to hit unforeseen bumps and bottlenecks as we refine, improve, and scale the system to meet everyone's needs.
"We have worked with Google to develop one-of-a-kind technology that can recognise videos based on a variety of factors. Video Identification is brand-new, cutting-edge stuff. Early tests with content companies have shown very promising results."