Megan, a robotics expert, is hired by a company to work on the project to create androids - robots who can physically pass for human under normal circumstances. Although there are a number of companies making "robots", none have so far produced one with true AI capability or such a human-like body. The problem Megan is to work on is the mental part. The first two prototypes of this line kept having trouble until they ceased functioning. The third prototype killed itself.
Megan finds the android, Ander, has been given such extensive and intense safeguards to protect humans that it is difficult for Ander to act or make choices that don't pose a potential conflict with the safeguards. She eases the safeguards and is able to make progress with Ander making choices and acquiring a more human behavior.
Another expert, Chandrarajan ("Raj") is also hired. He's very smart, but he's one of these eccentric geniuses who lives his own way. Others, especially supervisors, often have difficulty appreciating his behavior.
Ander is attracted to Megan. Megan and Raj are attracted to each other. This contributes to Ander tying up Raj and making advances on Megan.
The book gets us started with issues of robots / AI design and mental development after manufacture. This stimulation of thought on a topic is what most appeals to me. In the mid-book it transitions first into Megan being faced with uncertainty about accusations against and possible dangers from both Raj and Ander. This soon mixes with - then transitions to - a sort of chase-story / thriller-adventure theme. In this chase / thriller part, Ander takes both Megan and Raj with him on a quest to find other androids. Then they are pursued and kidnapped by criminals wanting to make money selling Ander.
Although the chase / thriller plot of the book can keep one's attention, such plots tend to be less idea-rich. And that part of the book is. That is disappointing to a reader such as myself. The story had begun asking questions about androids / AI before the transitions, but doesn't carry these through to as much speculation as I'd like. There is a quasi-answer in a sense in a story twist - but not in a particularly clear or extensive way.
Other than presenting the idea that safeguards intended to prevent robots from harming humans can be overly-strict leading to robots being unable to function, not much is examined about "laws of robotics". We are shown some of what a robot's "childhood" and "socialization" might be in an individual prototype environment, but the story does not extend to what this might be like in a mass production setting.
This may be hard SF in the sense of not being inconsistent with science. (There is one aspect later in the book which goes further towards the fringes of science.) But it's not the kind of hard SF that stays with the idea-rich story material.