Cisco Webex AI Agent AI Transparency Technical Note

Publisher

Cisco Systems, Inc. (Cisco Trust Portal)

URL

https://trustportal.cisco.com/c/dam/r/ctp/docs/white-paper/collaboration/ai-transparency-webex-ai-agent-technical-note.pdf

Published

2026-07-28

Updated

unknown

Accessed

2026-09-05

HTTP status

200

License

Cisco proprietary documentation, all rights reserved; no-redistribution; short excerpts and locators only

Claims citing this source

• The Webex AI Agent technical note contains no restriction on using its output for employment decisions and the word employment does not appear in it. fact in context
• Webex AI Agent supports autonomous agents driven by large language models for dialogue and state management and scripted agents using conventional machine learning for natural language understanding. fact in context
• In 5 of the 8 Webex AI Agent regions - Canada, Japan, Australia and New Zealand, Singapore and India - the Azure OpenAI inference column reads Global, which the table's own footnote defines as processing in any data centre worldwide where the models are hosted. fact in context
• Of the 8 Webex AI Agent regions, only London and Frankfurt are pinned outright (not marked Global) for Azure OpenAI inference, while the United States region reads US Regional plus Global. fact in context
• Cisco names 4 Azure OpenAI models behind autonomous Webex AI Agents: GPT-4o, GPT-4.1-mini, GPT-4.1 and a GPT-4.1 global variant used for variant generation and other features. fact in context
• Cisco states that the large language models powering autonomous AI agents come from Microsoft's Azure OpenAI Service. fact in context
• The Webex AI Agent note publishes a data locality table for 8 named regional deployments: produs1, prodeu1, prodeu2, prodca1, prodjp1, prodanz1, prodsg1 and prodin1. fact in context
• For the India Webex AI Agent region the internal model proxy is located in Sydney Australia while the cloud and speech components for the same region are in Mumbai and Central India. fact in context
• Cisco states that both autonomous and scripted Webex AI Agents support multilingual capabilities and publishes no list or count of supported languages anywhere in the note. fact in context
• Cisco states that Webex AI Agent outputs may contain inaccuracies or errors and strongly advises users to validate critical outputs before acting on them. fact in context
• Cisco states that payment card and protected health information are redacted by default for Webex AI Agent, and that customers may additionally mask names locations health data email addresses phone numbers and internet protocol addresses. fact in context
• Cisco states that customers may enable call recording for voice interactions handled by Webex AI Agents, and publishes no retention period for recordings or for agent transaction data, deferring retention to Offer Disclosures on the Trust Portal. fact in context
• Scripted Webex AI Agents use 1 of 2 natural language understanding engines, Swiftmatch or Mindmeld, both hosted in Webex Contact Center infrastructure so that training data and models stay within the customer's control. fact in context
• For the Singapore Webex AI Agent region the internal model proxy is located in Singapore while the Azure OpenAI inference for the same region is marked as processable in any data centre globally. fact in context
• The speech vendors behind Webex AI Agent vary by engine tier: Webex AI Pro 1.0 uses Azure for both speech to text and text to speech, while Webex AI Pro-US 1.0 and Webex AI Pro-EU 1.0 use Eleven Labs for text to speech and Deepgram for speech to text. fact in context
• 3 smaller Webex AI Agent components run on Cisco's own infrastructure: an interim response model on GPT-4.1-mini, a standalone query generator fine-tuned from facebook bart-large-xsum, and a turn detection ensemble fine-tuned from SmolLM2-135M-Instruct and bert-base-uncased. fact in context
• 3 of the 5 contact centre transparency notes carry a cover date of January 2025 and were 19 months old at the access date; the Topic Analysis note is dated March 2026 and the Webex AI Agent note carries no cover date in its body at all. fact in context

Cite this source record

APA

WarmTransfer. (2026, July 28). Cisco Webex AI Agent AI Transparency Technical Note. WarmTransfer. https://warmtransfer.net/knowledge/sources/cisco-trustportal-tn-webex-ai-agent

BibTeX

@misc{warmtransfer-cisco-trustportal-tn-webex-ai-agent,
  title  = {Cisco Webex AI Agent AI Transparency Technical Note},
  author = {{WarmTransfer}},
  year   = {2026},
  url    = {https://warmtransfer.net/knowledge/sources/cisco-trustportal-tn-webex-ai-agent},
  note   = {Cisco Systems, Inc. (Cisco Trust Portal), accessed 2026-09-05}
}