← All articles

Mikhail Polyakov

The ElliQ companion robot for older people in New York State: a new $2.1 million contract and how to read the figure "94% report feeling less lonely"

Level: BeginnerReading time: 5 min

Topics: Robots, AI

A small soft desktop device with a round glowing screen on a wooden stand by a window, beside a cup of tea, a knitted blanket and glasses on a stack of books

Short answer: the New York State Office for the Aging (NYSOFA) has signed a new three-year, $2.1 million agreement with Intuition Robotics to keep providing older residents with the AI companion ElliQ. The article gives impressive figures: for example, 94% of participants reported feeling less lonely. But the text does not show how these figures were obtained or how many people answered. Below: what was announced, what stands behind it and which questions to ask any seller of a similar solution.

What was announced

There is one source: a piece in The Robot Report. It says NYSOFA extended the partnership last week. I have not read the primary source (the announcement by NYSOFA or the company).

  • Who. NYSOFA, a state government office, and Intuition Robotics, the company that makes ElliQ. The company was founded in 2016 and is based in Ramat Gan, Israel.
  • What was agreed. A new three-year agreement worth $2.1 million. It continues the cooperation that began in 2022.
  • Who it is for. Older residents of the state who are at risk of loneliness and social isolation.
  • Scale. Since the launch in 2022, hundreds of older residents of the state have received ElliQ. In 2025 alone, more than 3,500 people applied. The text does not say how many devices the $2.1 million covers.
  • Other states. The company has similar programmes in Wisconsin, Florida and Washington. In Washington the programme is aimed at Medicaid recipients.

Which figures are given

The text calls them "newly reported program data from 2026":

  • an average of 48 interactions with ElliQ per day, compared with 29 in reports from previous years;
  • 94% of participants report feeling less lonely, 97% report improved overall wellbeing, 88% say ElliQ has made their daily life better (in the text: "positive difference in their daily lives"), and 79% feel more connected to the world;
  • every month 88% of participants have a conversation with ElliQ, 92% regularly share updates about their wellbeing, and 81% take part in physical activity that it encourages.

The director of NYSOFA, Greg Olsen, called ElliQ an important tool for fighting the "loneliness epidemic". The company's co-founder and CEO, Dor Skuler, described the results as "measurable, record-breaking engagement results".

What remains unclear about these figures

The text does not show:

  1. Who measured, and how. The figures are called programme data, but it does not say who collected them, whether it was a survey, or how the questions were asked. "94% report feeling less lonely" sounds like participants' self-assessment.
  2. How many people answered. A percentage without the number of people and without the share of all participants who answered says little.
  3. What it was compared with. The text says nothing about a group without ElliQ, so it is impossible to say how much of the change is due to the device and not to something else.
  4. Who reported the figures. They are given without naming whoever made the measurement; next to them are the words of the head of the selling company.

This does not mean the figures are wrong. It means that, from the text, one cannot judge how far they can be relied on.

What to take from this

If you are choosing an AI helper for older relatives, for people your organisation looks after or for clients, you can ask the seller: what the effect was measured with and how, who answered the questions, how many people, whether there was a comparison group, who pays for the device and what happens to the data when the contract ends. Who does it: the buyer, before signing the contract. How to check: ask the seller for a written description of the method. If you get one, assess it against the five questions above: who answered, how many people, what was compared. If the seller is unwilling or unable to answer, treat the figures as the seller's own claim and do not rely on them when choosing. A trial with a small group before buying for everyone is possible if the seller offers one: decide in advance what you will judge success by, for example participants' answers to the same questions before and after.

The text also recalls that NYSOFA formed its first public-private partnership in 2018, after a report by AARP and the US Centers for Medicare and Medicaid Services (CMS) on the effect of loneliness and isolation on older adults and on Medicare spending. As retold, the report compares the consequences of loneliness and isolation to smoking almost a pack of cigarettes a day and puts Medicare's cost of treating them at more than $7 billion. I have not read the report itself, so this is a retelling of a retelling.

Summary

NYSOFA has extended its programme with ElliQ for three years at $2.1 million, after hundreds of older residents have received the device since 2022 and there were more than 3,500 applications in 2025. The figures on reduced loneliness are given in the text without a description of how they were measured, so they should be read as programme data and not as an independent check. Little is known about the contract itself: how many devices it covers and what is included in the price are not stated in the source.

See also another piece in this section: "Boring" robots versus spectacular videos.

I work on AI agents and automation. If you would like to look at my projects or discuss your own task, visit my portfolio.

Sources