Image recognition is a key subset of artificial intelligence that, as the name implies, can tell what, and increasingly who, is in a photograph. And, it is a key battleground for Microsoft (MSFT), Google (GOOG), Facebook, and Amazon.

Related: Microsoft Takes Aim at Google, Facebook With New Recognition Tool

The latest is that Amazon (AMZN) Rekognition software, which debuted last fall, has been updated so it can identify hundreds of thousands of famous or prominent people including politicians, athletes, actors, and business and media leaders from around the world, according to a company blog post.

Get Data Sheet, Fortune’s daily technology newsletter.

Developers can use this tool to add image recognition to their own software. The first version could determine if a picture was of a human face and, within a high probability, identify that person’s gender, whether he or she was wearing glasses, was smiling, and other key characteristics. What it did not do is specifically tell you who that person may be. Amazon has now raised the stakes so the software can identify the individual, provided that person is well known.

Related: Amazon Snaps Up Image Recognition Startup

Developers can use Rekognition to analyze and index their own photo archives and to develop image search tools, according to the blog.

Image recognition could theoretically be used, in an interactive retail display to tell if someone standing nearby is a man or a woman and adjust the sales pitch accordingly. But as the technology gets better at identifying actual individuals, who knows where it will end up. Facebook (FB), for example, can now identify people in pictures even if they are not tagged with the person’s name.

Related: Google Says Its Face Recognition Software is the Best

Companies that deliver smart image recognition can make money applying it to sales, marketing, and security jobs. But some people worry whether software is getting too smart for our own good and see these advances as more an intrusion than progress.