For companies like Google and Microsoft, AI is an important part of their future, enhancing existing products and expanding revenue sources. However, as recent financial reports reveal, the two companies also acknowledge that AI, especially biased AI that makes wrong decisions, may damage their brands and businesses.
according toForeign media news
From 10-K submitted by Alphabet last week:
From Microsoft's 10-K submission last August:
But these companies only point out that AI is now a potential factor. After all, the two companies have been developing AI products for many years, starting from the Google autopilot car program in 2009 and the long-term cooperation between Microsoft and Cortana. This technology has ample opportunities to cause brand damage, and in some cases already exists. Remember when Microsoft's Tay Chat Robot was launched on Twitter and started promoting racist nonsense in less than a day? Years later, it is still often cited as an example of AI errors.
However, you can also say that public awareness of AI and its potential adverse effects have increased significantly over the past year. The scandals of Google's secret work with the Pentagon under Maven, Amazon's biased face recognition software, and Facebook's incompetence with the algorithms used to analyze the Cambridge scandal have brought AI problems, which are seriously enforced, into the spotlight.
Microsoft and Google have increased risk prevention: Microsoft, for example, is discussing the need to regulate facial recognition software to prevent potential hazards, and Google has begun to communicate with policymakers and academics on AI governance.