Debated to death and some have interesting responses on the subject.
An automated response is not sentient as it is a product of artificial electrical/chemical sensors that operate as programmed like a shopping centre's programmmed auto sliding doors or when water turns to ice under freezing temperatures.
So....I have heard/read from various sources coming out with these propositions:
1. Such responses are lacking in the criteria of having conscious thought or intention as how the Pali texts (speaks of citta (consciousness) and cetana (volition). The 'Buddha Nature' is found only in sentient beings where 'Bodhicitta', the important element of one's cultivation (as how Mahayana/Vajrayana defines it), is found and capable of being made manifest.
2. So, unless a 'being' has the 5 aggregates (the formless ones lacking form, so it will be down to 4), technically speaking, no, it is an 'automaton', living yes but not sentient, no matter what smiley or Barbie or Dr Spock's face that one paints on it, it is not equaled to the above 'definition'.
3. Another one is if a being is not capable of experiencing (as opposed to being indifferent) dukkha, sentience is ruled out.
4. It is said that because of sentience, we are somewhat motivated to cultivate, understanding the imminence of transcending Samsara and suffering, hence we cultivate the Buddha Dharma, the source of our liberation.
5. No where in the Buddha Dharma's 'cosmology', as taught in the 3 Traditions, does it mention that one takes birth as a plant or as Data (in Star Trek). I cannot imagine how vegetarians/vegans would cringe if they were ever told that plants were sentient!
6. Despite the 'un-sentient' quality of plants and programmed machinations, we nevertheless show respect and conservation of them as part of our Dharma practice and cultivation. One example is found here
An interesting read though an old experiment is the Turing Test
and its 'refutation' here
Is a 'sentient android' possible?
Namo Amitabha Buddha!