Raymond Perrault

Confidence 0.85 · 3 sources · last confirmed 2026-05-28

Distinguished Computer Scientist Emeritus at SRI International (Director of the SRI AI Center, 1987–2017). One of the longest-serving figures on the AI Index Steering Committee — chair of the 2025 edition and co-chair (with Yolanda Gil) of the 2026 edition; the 2026 Index’s headline “The data does not point in a single direction” framing comes from the Co-chairs’ message authored jointly with Gil.

Research grounded in natural language processing and speech act theory. From 2002 to 2009 was co-Principal Investigator on the CALO Project — DARPA-funded multi-institutional effort to build a personal-assistant agent that learns through interaction. CALO is the technological lineage from which Siri (and several other commercial/military assistants) was spun out, making Perrault a meaningful lineage figure for the wiki’s ai-agents and agent-harness threads even though his recent surface in the Index is governance-coloured rather than systems-coloured.

Appears in this wiki via

• 2026-04-28-ai-index-report-2025 — co-author and chair, 8th edition.
• 2026-04-30-ai-index-report-2026 — co-author and co-chair, 9th edition.
• 2026-05-27-sajadieh-stanford-hai-inside-the-2026-ai-index-report — panel principal at the 27 May 2026 Stanford HAI talk-track; first-person founder-account of the AI Index’s facts-only editorial discipline; names the acceptability-threshold research gap (“how well does it have to work to be acceptable in a legal / medical / finance system?”) and discloses the joint World-Bank / IMF global-AI-adoption measurement initiative with Yolanda Gil.