Verified data from 2022–2026

The State of Enterprise AI Assistants

From pilots to production: how the world's largest organisations are deploying AI assistants at scale

0+

Walmart associates using AI weekly

[4, 1]
0

JPMorgan employees on LLM Suite

[76, 77]
0+

Bank of America Erica interactions

[78]

Research by Faisal Shariff

0:00 / 0:00

A curated analysis of notable enterprise AI deployments across global retail, agentic platforms, and cross-industry implementations. This research highlights significant publicly disclosed deployments and is not intended as an exhaustive market survey.

Scroll to explore

The Retail AI Landscape

Every major retailer has moved beyond pilot. Here's who launched what, when, and at what scale.

Walmart

Ask Sam

Powered by In-house + Microsoft

ProductionColleague-facing
Launched

2023-08

Scale

900K+ weekly users[4, 1]

Target

Store Companion

Powered by In-house

ProductionColleague-facing
Launched

2024-03

Scale

~2,000 US stores[8, 10]

Kroger

Instacart AI Search

Powered by OpenAI (ChatGPT)

ProductionCustomer-facing
Launched

2025-11

Scale

Agentic shopping[12, 11]

Carrefour

Hopla

Powered by OpenAI

ProductionCustomer-facing
Launched

2023-06

Scale

carrefour.fr users[14, 16]

Ahold Delhaize

Steijn (Albert Heijn)

Powered by In-house

ProductionCustomer-facing
Launched

2024

Scale

Millions of app users[17]

Amazon

Cedric (Internal)

Powered by In-house (Bedrock)

ProductionColleague-facing
Launched

2024-09

Scale

Internal deployment[19, 20]

Tesco

Store & Colleague Intelligence

New partnership with Mistral AI

ProductionMixed
Launched

2025-12

Scale

Enterprise-wide[22]

Sainsbury's

AI-Guided Colleague Tools

Powered by Microsoft Azure

ProductionMixed
Launched

2024-05

Scale

All UK stores targeted[27, 25]

Marks & Spencer

Store Intelligence

Powered by SymphonyAI

ProductionColleague-facing
Launched

2023-09

Scale

500+ stores[28, 29]

Aldi

GenAI Content Platform

Powered by Digital Wave Technology

FoundationMixed
Launched

2025-03

Scale

US digital channels[31]

Lidl

AI Operations Suite

Powered by In-house + Cognigy

FoundationMixed
Launched

2024

Scale

European operations[111]

AI Enterprise Research · Retail Intelligence

Global Grocery AI Landscape

Deployment scale vs. capability sophistication across global grocery retail. Hover for details.

United States
United Kingdom
European Union
You are here
Capability · Basic → Agentic
Deployment Scale · Narrow → Enterprise-wide
Experimenting
Early pilots, limited scale
Scaling Ops
Broad deployment, basic AI
Chat Pioneers
Advanced AI, narrow reach
AI-First Retail
Advanced AI at scale
🇺🇸
Walmart
🇺🇸
Kroger
🇺🇸
Target
🇺🇸
Amazon
🇳🇱
Ahold Delhaize
🇫🇷
Carrefour
🇬🇧
Tesco
🇬🇧
Sainsbury’s
🇬🇧
Marks & Spencer
🇩🇪
Aldi
🇩🇪
Lidl
Last updated: February 2026 · Source: Atomic Theory Consulting competitive intelligence

The Search-to-Chat Transformation

Search is fragmenting. Chat is rising. The data tells the story.

Google AI Overviews

(formerly SGE (Search Generative Experience))

0%+

of searches now end without a click

[35]
0%

of marketers report reduced organic traffic from SGE

[34]
0%

potential CTR reduction for long-tail queries

[36]

Ask whatever’s on your mind

Google AI Overviews[39]

Microsoft Copilot

0M+

monthly active users

[40, 43]
0%

of Fortune 500 adopted by FY25 Q1

[41]
0%

projected 3-year ROI for SMBs

[45]
0 hrs

saved per user per month

[44, 45]

Top User Intents

Searching51.6%
Doing34.6%
Expressing13.8%

Analysis of 37.5 million conversations (Jan–Sep 2025): desktop dominated by 'Work and Career' (8am–5pm), mobile dominated by 'Health and Fitness' at all hours

50% of tech leaders unsure $30/user/month is justified. Jefferies projects $11B revenue by CY2026.

$30/user/month

Google Gemini

0M

monthly active users by Q2 2025

[49, 42]
0M

enterprise users of Gemini Pro globally

[50]
0%

of Gemini usage from enterprise users

[50]
0%

of Fortune 500 using Gemini in at least one department

[42]

Workspace Integration

2.3 billion document interactions in H1 2025; Gemini in 73% of enterprise accounts

Bundled into Workspace Business/Enterprise tiers since early 2025

Agentic Capabilities in Production

The shift from 'assistants that inform' to 'agents that act' accelerated dramatically in 2024–2025.

1
Level 1 · Inform

Answer questions, summarise documents, search knowledge bases

Target Store CompanionAmazon CedricCarrefour Hopla
2
Level 2 · Create

Generate tickets, records, documents, presentations

JPMorgan LLM Suite (5-page deck in 30 sec)Microsoft Copilot Agent Mode in Word
3
Level 3 · Update

Modify data, change statuses, process approvals

Salesforce AgentforceServiceNow AI Agents
4
Level 4 · Execute

Book meetings, submit forms, trigger workflows, resolve tickets

Zendesk AI Agents (80% resolution)Intercom Fin (66% avg resolution)DHL HappyRobot
5
Level 5 · Orchestrate

Multi-step workflows across multiple systems with autonomous decision-making

ServiceNow AI Agent Fabric (cross-system)Microsoft Agent 365Stellantis supply chain agent
Standout Case Study

ServiceNow × Orica

IT Service Desk deflection

Before0%
After0%

Doubled fully resolved cases without human intervention[56]

Salesforce Agentforce

GA 2024-10-25

Shift from Einstein (recommendations) to Agentforce (autonomous execution)[52, 51, 53]

Customer Results

1-800-Accountant

Autonomous chat resolution during tax week

70% resolved without humans[54]

Key Metrics

  • 18,500 deals closed across 39 countries
  • Agentforce accounts in production up 70% QoQ
  • ARR up 330% YoY, surpassing $500M

Notable Demo

UK Police deployed 'Bobby' agent for non-emergency calls across multiple departments, covering 90+ topics, delivering 20% reduction in human agent demand.

Marc Benioff's vision: one billion agents by end of 2025

Capabilities

Retrieves data, reasons through options, builds action plans, executes them

ServiceNow AI Agents

GA 2025-05

AI Agent Fabric for cross-system orchestration across ServiceNow, Microsoft, Google, and Box[56, 57, 58]

Customer Results

Canada Life

Catalog creation dev time

200% reduction[56]

Key Metrics

  • 18% reduction in escalated incidents
  • 33% reduction in incident resolution time
  • 15–25 minutes reduction in major incident resolution time

Notable Demo

Stellantis supply chain: AI agent detected 25% battery cost increase, recommended alternate supplier, conducted capability analysis — all from single human alert

Microsoft Copilot Agent Mode

GA 2025-11

Agents plan, execute, and validate multi-step tasks directly in Office apps[59, 60, 61]

Capabilities

Creating documents, financial models, table reshaping, presentations from conversational prompts

Grounding

Microsoft Graph for permission-checked retrieval with transparent citations

Zendesk AI Agents

GA 2024

[62, 63, 64]

0%of support issues resolvable without humans[62]
0+AI customers[64]
0Btickets processed per year across ~20,000 customers[64]

CSAT up 5–10 points; 35% reduced ticket volume; 20% reduced response time; day-one automation of 30% from knowledge sources alone

Intercom Fin

GA 2024

[65, 66, 67]

0%average resolution rate (increasing ~1%/month)[66]
$0.99 per resolution

Pre-Fin AI, teams reported 11–30% of support volume resolved by AI bots

Cross-Industry Scoreboard

AI assistants at scale across banking, healthcare, telecoms, and manufacturing.

JPMorgan

LLM Suite

Award

Scale: 250K employees, ~half daily

5-page IB deck in 30 seconds[76, 77, 83]

Tech: OpenAI + Anthropic models (multi-model strategy)

American Banker 2025 Innovation of the Year

Deep Dive

Launched as secure corporate ChatGPT in 2023, scaled to 250K in 8 months. Now layering agentic AI for regulatory reports. Staff may fall 10% over 5 years.

Bank of America

Erica

Scale: 50M users, 58M interactions/month

3B+ total interactions; 98% find info needed[78, 79]

Tech: In-house, launched 2018

1.7 billion proactive insights delivered. Longest-running major bank AI assistant.

Barclays

Colleague AI Agent

Scale: 15K pilot → 100K global rollout

Travel booking, HR policies, compliance search[80, 82]

Tech: Microsoft Copilot integration

HSBC

Multiple AI tools

Scale: 600+ AI use cases; 20K developers with coding assistants

15% coding efficiency gain; 3M client interactions/yr via GenAI (88% ease)[81, 82]

Tech: Google partnership; LLM productivity tool for all colleagues

Group-wide mandatory AI training plus 'AI Academy' for beginner-to-advanced skills

Goldman Sachs

GS AI Assistant

Scale: Firmwide June 2025

Largest AI talent share growth in banking index[82, 80]

Tech: In-house

Morgan Stanley

AI @ MS Assistant + AskResearchGPT

Scale: 16K advisors (WM, 98% adoption) + Institutional Securities

98% advisor adoption; document access 20% → 80%[84, 85, 87, 90, 92, 95]

Tech: OpenAI GPT-4 (enterprise arrangement)

Two flagship AI assistants: AI @ MS Assistant serves 16,000 Wealth Management advisors with instant access to proprietary research and policy content — 98% of advisor teams are active users. AskResearchGPT serves Institutional Securities, synthesising 70,000+ research reports annually with one-click email-draft export and inline citations. Both grounded in proprietary content only, with human-in-the-loop for all client-facing outputs. Bank aims to double client assets to $10 trillion partly through AI-freed advisor capacity.

Voice & Multimodal AI Market

Market Size (2024)$0B
Projected (2034)$0B
CAGR
34.8%

Retail voice commerce CAGR: 31.5% through 2030 through 2030[74, 75]

The ROI Evidence Wall

Published, cited figures from production deployments. No vendor hype.

Walmart

Shift planning time

Before90 min
After0

minutes

Walmart Corporate[4]

Walmart

CX scores

Before
After0

basis points

Walmart Earnings Call FY 2024[7]

ServiceNow (Orica)

IT service desk deflection

Before18%
After0%

ServiceNow Knowledge '25[56]

Vodafone (Portugal)

First-time resolution

Before15%
After0%

Vodafone / IOT World Today[72]

Vodafone (Portugal)

Net Promoter Score

Before14
After0

NPS

Vodafone / IOT World Today[72]

NHS Copilot Trial

Time saved per staff member

Before
After0

min/day

NHS England, 30K users[96, 97]

NHS (projected)

Monthly hours saved at scale

Before
After0

hours/month

NHS England[98, 96]

Bank of America

Users finding info needed via Erica

Before
After0%

BoA Press Release[78]

Zendesk

Ticket volume reduction

Before
After0

% reduction

Zendesk[62, 64]

JPMorgan

IB deck generation time

BeforeHours
After30 seconds

CNBC[76, 83]

NHS (Appointments)

Missed appointments (DNAs)

Before
After0

% reduction

NHS England[99]

Microsoft Copilot (SMB)

3-year ROI

Before
After0

% ROI

Forrester TEI Study[45]

Microsoft Copilot

Time saved per user

Before
After0

hours/month

Forrester[44, 45]

Bot Sprawl & Federation

The average enterprise has 12+ separate AI entry points. Fragmentation kills adoption.

12+

separate AI entry points across vendor-bundled bots in the average US enterprise[105, 106]

Each requires separate authentication and loses context at every hand-off

Adoption Gap

Integrated>60% adoption rates
Fragmented<15% adoption rates

Gartner Prediction

By 2026, more than 50% of enterprises will have failed to implement an AI governance framework[105, 106]

Federation Solutions

Meta-Bot Architecture

Single entry point that ingests user intent, determines which backend system should handle it, and routes accordingly[105]

ServiceNow AI Agent Fabric

Agents communicate and hand off tasks across ServiceNow, Microsoft Azure, Google, and Box[109, 56]

Microsoft Agent 365

Control plane for managing and securing agents with enterprise-wide visibility[60]

Academic: Intent-Based Orchestration

Centralised intent analysers, orchestration engines, and reasoning systems creating a single access point with domain collaboration[107]

Governance Best Practices

1

Central agent registry tracking all agents, owners, purposes, permissions

2

Standardised templates and approved configurations to reduce duplication

3

Single entry point for users with intelligent routing to backend systems

4

Human-in-the-loop for high-stakes actions with approval workflows

5

Cross-team coordination to prevent shadow AI and duplicate development

What Failed (and Why It Matters)

The lessons that don't make it into vendor pitch decks.

McDonald's

AI Drive-Thru Cancelled

High Severity
June 2024

IBM partnership from 2021, tested at 100+ US locations. Accuracy was only 80% — below the 95% target.

TikTok videos showed AI adding 260 McNuggets ($211), bacon to ice cream, and multiple sundaes instead of a vanilla ice cream

Outcome

Instructed franchisees to remove technology by 26 July 2024[75]

Lesson Learned

Voice AI in noisy, high-variance environments requires far higher accuracy than currently achievable

Air Canada

Chatbot Legal Liability Ruling

High Severity
February 2024

Chatbot provided incorrect bereavement fare information to a grieving passenger.

Air Canada argued the chatbot was a 'separate legal entity' — the BC Civil Resolution Tribunal called this a 'remarkable submission' and rejected it

Outcome

Ordered to pay $812.02 in damages. Chatbot removed from website.[106]

Lesson Learned

Companies are legally responsible for all information on their websites, regardless of whether it comes from a static page or a chatbot

Google

AI Overviews Launch Errors

Medium Severity
May 2024

Within days of US rollout, AI Overviews produced viral errors from unvetted sources.

Suggested non-toxic glue for pizza (from 11-year-old Reddit joke) and eating rocks (from The Onion)

Outcome

Google acknowledged failures and made adjustments[37, 38]

Lesson Learned

AI-generated content drawing from unvetted sources at scale creates massive reputational risk

Systemic Challenges

$67.4 billion

Hallucination Costs

Estimated enterprise losses globally in 2024. 82% of AI bugs stem from hallucinations — 'silent misinformation' that looks perfectly normal.[106]

$14,200 annually for mitigation

50%

Copilot ROI Uncertainty

of tech leaders unsure $30/user/month is justified. Companies in staged deployment, not full rollout.[48, 47]

Shadow AI Risk

Departments independently adopting ungoverned AI tools when official channels are too slow, creating compliance and data security exposures.[106, 108]

Sources & Methodology

Every claim in this report is traceable to a named source.

This research draws on primary sources (company announcements, earnings calls, corporate blogs), analyst reports (Forrester, Gartner), trade press, and vendor case studies. All data points are from 2022–2025. Claims are cited to their original source.