PHILIPPINE companies must upgrade detection and infiltration methods to protect consumers against artificial intelligence-(AI) generated deepfakes, according to mobile app security solutions provider Appdome.
“The reality is the line between what’s real and not real is not that bright anymore, and so it’s very hard for the average consumer to know when something’s fake versus when it’s not,” Tom Tovar, co-creator and CEO of Appdome, said in an interview on Money Talks with Cathy Yang on One News on Feb. 20.
“I really think the responsibility is on the brands and the people who provide the applications to consumers to step forward and put (detection technology) in the application to keep users safe (by detecting) those methods of infiltration, manipulation, and substitution that AI needs to get fake content and fake images into your user experience.”
According to Mr. Tovar, the Philippines is considered a “mobile-first” economy, with its average use of mobile apps higher than the global average.
About 54.7% of Filipino consumers said they use mobile applications more than the web, according to Appdome’s Consumer Survey of Mobile App Security.
However, high dependency on mobile apps makes it vulnerable AI-driven fraud, including deepfakes, Mr. Tovar said.
“Deepfakes are one of the top attack vectors that have emerged in the last two to three months,” he added, noting that mobile-driven countries must prioritize authentication and user login.
With AI, deepfakes have evolved more sophisticated manipulation techniques such as face, image, and voice cloning. This also increases the risk of disinformation ahead of this year’s midterm elections.
“You might think you’re getting a call from your government official when reality, it’s not (them). It might sound like them, it might appear like them,” Mr. Tovar said. “So, AI has the potential of also being a mechanism for disinformation as much as it is a wonder to use in a legitimate way.” — Beatriz Marie D. Cruz