What’s changing—and why it matters Explore developments across AI platforms and industries, where organizations are putting them to work, and what they mean for business decisions.
WhyteOps research powered by AI. Interpreted for business.
Updated September 29, 2026 Coverage: July–September 2026
7
July 2026
Dated developments
5
August 2026
Dated updates
16
September to date
Updates through September 29
5
Earlier customer reports
Context—not new July activity
Counts describe selected public evidence, not growth rates. The research set includes AI developments and relevant digital-data infrastructure. Eleven organization labels are represented.
DEVELOPMENTS OVER TIME
What changed, and who reported it?
Compare dated developments by company and sector. Open any record for its evidence and limitations.
33 records: 28 from July to now and five earlier background reports.
MONTHLY REVIEW
July 2026
Capabilities and controls
Platform capabilities and controls were already taking shape. AT&T reported customer-care summaries, while FDA selected a digital-health pilot participant.
GooglePreview
OpenAIIncident disclosure
FDAPilot selection
AnthropicRelease
AT&TReported availability
GoogleRelease bundle
AnthropicIncident disclosure
No matching update in this research set. This does not mean no progress occurred.
7 records shown
FOLLOW-UP
August 2026
Specialized offerings and expansion plans
The selected updates add specialized offerings, a retail data-platform expansion plan and reported security remediation.
WalmartPlanned for 2027
GoogleIndustry offering
GoogleIndustry offering
AnthropicResearch preview
AnthropicControl update
No matching update in this research set. This does not mean no progress occurred.
5 records shown
MONTH-TO-DATE
September to date
Through September 29
New model and connected-app releases, a Copilot roadmap, security operations, and commerce experiments extend the evidence. Availability, planned rollouts and reported use remain distinct.
NASA / MicrosoftReported operational use + prototype
No matching update in this research set. This does not mean no progress occurred.
16 records shown
The pre-July customer reports are shown below as background. They are not new July deployments.
WhyteOps interpretation
Across this period, the evidence connects expanding technical capabilities with specific workflows and the controls needed to use them. The practical question is not simply “who announced AI?” but “what changed, at what stage, and what should we evaluate next?”
WhyteOps Trend Index
What is changing, where it is used, and why it matters.
Trend / theme
Observed development
Customer use
WhyteOps relevance
Evidence
Bounded work is moving from answers into actions.
AT&T reports service-update agents.
Map approvals and exceptions.
8 records
Runtime, context and tools surround the model.
FDA reports connected-data integration underway.
Assess the operating environment.
4 records
Access controls and monitoring accompany execution.
AT&T describes checkpoints and action logs.
Test controls inside the workflow.
13 records
Reliable information makes AI useful at work.
Customer examples across four sectors.
Improve ownership and source quality.
13 records
Specialized agents hand work to each other.
AT&T reports coordinated service agents.
Define handoff ownership and recovery.
2 records
Theme references overlap and must not be added together. Momentum, adoption percentages and overall scores are not established. Select a theme to examine its sources.
What was already in place before July?
Historical customer reporting gives our July snapshot context. These five records are not new activity during the tracking period.
healthcare
Kaiser Permanente
2.5M+
encounters in the historical evaluation
Operational use
AI-assisted clinical documentation
7,260 physicians used AI scribes in 2,576,627 encounters during October 2023–December 2024. Physicians could review and edit notes.
Published 2025-04-01. Encounters are not unique patients. Observational findings are not proof of causation or autonomous clinical decisions.
retail
Walmart
90 → 30
minutes: early shift-planning estimate
Operational use + pilots
AI-assisted shift planning
Early manager estimates put planning time at 30 minutes instead of 90 for an overnight-stocking workflow; other shifts were in pilot.
Published 2025-06-24. An early estimate, not a controlled study. Workforce scope is not an active-user count; planned GenAI upgrades are separate.
telecom
AT&T
Production
reported service-update workflow
Reported production
Agents handling service updates
AT&T describes agents processing service-update requests and synchronizing systems, with human checkpoints and action logs.
Published 2025-11-13. Benefits are qualitative. The separate network-repair scenario in the article is illustrative, not counted as a confirmed deployment.
Federal government
FDA
June 2025
launch announcement
Reported rollout
Elsa agency-wide launch
FDA reported launching its internal assistant and using it in review and inspection-related work.
Published 2025-06-02. Availability is not universal active use, validated accuracy or a government-wide adoption measure.
Federal government
FDA
May 2026
upgrade announcement
Upgrade; integration underway
Elsa 4.0 and connected data
FDA reports an assistant upgrade, data consolidation into HALO, and the beginning of integration for connected workflows.
Published 2026-05-06. Integration was underway. No comparable productivity outcome is supplied against the 2025 launch.
No matching pre-July customer report for these filters. Use “Before July · background” or reset filters to explore the starting context.
Development over time: FDA
An internal assistant launches
FDA reports Elsa’s rollout and use in review-related work.
Connected data and custom agents
FDA reports an upgrade and begins integrating its consolidated HALO data platform.
The story
A progression from an assistant toward connected organizational workflows. This shows capability development—not measured productivity growth.
Latest evidence publications
Historical reports retain their original publication dates.
Anthropic
NASA / Microsoft
Microsoft
Anthropic
Google
T-Mobile
Google
Amazon
Anthropic
OpenAI
Google
OpenAI
Walmart
OpenAI
Kaiser Permanente
Anthropic
Anthropic
Anthropic
Google
Google
Walmart
Anthropic
Google
AT&T
Anthropic
FDA
Google
OpenAI
FDA
AT&T
Walmart
FDA
Kaiser Permanente
No matching public update in this research set. That is a coverage gap, not evidence of no progress.
