Memindex

August 24, 2026

Agent memory is now a governance problem, not a cache

Google's Memory Bank GA — TTL, revisions, IAM-scoped reads, residency rules — shows what a production write path costs teams building their own.

The interesting thing about Google's Gemini Enterprise Agent Platform going GA on July 29 isn't the headline feature. It's what shipped around the memory store. Per Google Cloud's documentation, Memory Bank offers TTL-based automatic expiration, memory revisions that let you inspect how a memory transformed as new information was ingested, similarity search scoped to a specific identity, and IAM conditions restricting which principals can read or write which memory scopes. The docs also name a failure mode explicitly — "cross-border contamination," where an agent in one jurisdiction writes to or reads from a Memory Bank instance in another — and recommend region-specific agent identities or service accounts as the control.

That list is a decent spec for anyone building their own memory layer, because each item is a bug report from production.

The write path has different failure modes than retrieval

RAG has no write side. You index documents someone else owns, and staleness is their problem. Memory inverts this: your system generates derived artifacts, and nobody upstream is maintaining them.

Two consequences. First, staleness is silent. A retrieved document has a visible timestamp and provenance; an extracted memory is a confident sentence with no natural expiry. TTL is a blunt instrument — most teams will want per-scope policies, short for state, long for stable preferences — but a default of "forever" is a bad default. Second, memories launder permissions. A fact extracted from a doc a user could read in March survives their offboarding in August unless reads are re-checked against current identity. Scoping similarity search to identity at query time, rather than filtering after retrieval, is the version of this that doesn't leak through ranked results.

Revision history matters for a less obvious reason: debugging. When an agent asserts something wrong, you need to know whether retrieval failed or whether the memory itself drifted across consolidation passes. Without a revision log, those look identical from the trace.

Runtime duration forces the tiering question

The same release pairs Memory Bank with an Agent Runtime supporting asynchronous runs up to seven days, and keeps Sessions (state within one interaction) separate from Memory Bank (recall across sessions). Google Cloud cited AT&T, Best Buy, Commerzbank, Liberty Global and WellSky among customers using or evaluating parts of the platform. Multi-day horizons make conflating the two tiers expensive — you either replay a week of session state into context or you promote the wrong things into long-term storage.

Two nearby dates for builders: OpenAI's stateful Assistants API is reported to sunset August 26, replaced by the stateless Responses API alongside a Conversations API for server-managed history — the same split between ephemeral turn state and durable history.

And don't shop on benchmarks. Mem0's own writeup notes its 92.5 LoCoMo figure and Zep's 94.7 claim use different models and evaluation setups; BEAM was built specifically so no current architecture saturates it. Evaluate on your own transcripts.

Sources

  1. [1] Context Engineering - LLM Memory and Retrieval for AI Agents | Weaviate
  2. [2] Memory as Action: Autonomous Context Curation for Long-Horizon Agentic Tasks
  3. [3] Look Back to Reason Forward: Revisitable Memory for Long-Context LLM Agents
  4. [4] Context Engineering: Memory and Temporal Context
  5. [5] InfoMem: Training Long-Context Memory Agents with Answer-Conditioned Information Gain
  6. [6] SWE-EVO: Benchmarking Coding Agents in Long-Horizon Software Evolution Scenarios
  7. [7] Memory for AI Agents: A New Paradigm of Context Engineering - The New Stack
  8. [8] From Theory to Practice: Context Engineering and Memory for LLM Agents | by Jovan Njegic | Medium
  9. [9] CoDA: A Context-Decoupled Hierarchical Agent with Reinforcement Learning
  10. [10] Memory in the Age of AI Agents
  11. [11] The definitive guide to AI‑based enterprise search for 2026
  12. [12] Enterprise AI search in 2026: What you need to know | Dust Blog
  13. [13] AI Enterprise Search Tools and Features for 2026 | Slack
  14. [14] Best Enterprise Search for Engineering Teams (2026)
  15. [15] The AI Enterprise Search Guide for IT and Knowledge Leaders
  16. [16] Enterprise Search in 2026: Why It Finally Works (and What Changed) | Atolio
  17. [17] Enterprise Search Options 2026
  18. [18] Enterprise AI Search: AI-Powered Search for 2026
  19. [19] RAG, Agentic RAG, and AI Memory - by Avi Chawla
  20. [20] Agent Memory Vs RAG: What Breaks At Scale 2026 (Analyzed)
  21. [21] AI Agent Memory 2026: Progress Benchmark Report Evaluations
  22. [22] RAG is Dead. Long Live Agent Memory - Chris Latimer - YouTube
  23. [23] Knowledge and Memory Beyond RAG: Why 2026 Agents Need a Write Path, Not Just a Retriever | by Micheal Lanham | Apr, 2026 | Medium
  24. [24] [2602.02007] Beyond RAG for Agent Memory: Retrieval by Decoupling and Aggregation
  25. [25] AI Memory Systems Statistics You Need to Know in 2026 (60+ Sourced Stats)
  26. [26] Machine Learning Pills
  27. [27] News from generation RAG - Dive deep into the transformative world of AI Retrieval Augmented Generation (RAG) technologies
  28. [28] LLMs with retrieval-augmented generation: Good or bad for privacy compliance? | IAPP
  29. [29] RAG is DEAD!. Retrieval-Augmented Generation ruled… | by Reliable Data Engineering | Medium
  30. [30] A Systematic Review of Key Retrieval-Augmented Generation (RAG) Systems:Progress, Gaps, and Future Directions
  31. [31] Beyond Retrieval: The Expanding Universe of Augmented Generation in AI
  32. [32] Enhancing LLM Performance with Retrieval-Augmented Generation
  33. [33] 1 Retrieval-Augmented Generation for Large Language Models: A Survey
  34. [34] Google AI Agents Get Seven-Day Runtime and Memory Bank — Enterprise DNA
  35. [35] AI Agents News — Week of August 24, 2026 (Daily Updates)
  36. [36] AI Agents News | August, 2026 (STARTUP EDITION)
  37. [37] Gemini Enterprise Agent Platform release notes | Google Cloud Documentation
  38. [38] What’s New in Oracle AI Agent Memory: Custom Extraction, Hybrid Search, and More Control | developers
  39. [39] Gemini Enterprise Agent Platform Updates by Google - July 2026 - Releasebot
  40. [40] News Desk: Knowledge Graphs, Agent Memory — GroundingNodes
  41. [41] The State of AI Agent Memory in 2026: What the Research Actually Shows - DEV Community
  42. [42] Gemini Enterprise AI Platform | Google Cloud
  43. [43] Google Cloud expands Gemini enterprise agent controls
  44. [44] Agent Platform Memory Bank | Gemini Enterprise Agent Platform | Google Cloud Documentation
  45. [45] Introducing Gemini Enterprise Agent Platform | Google Cloud Blog
  46. [46] Agent Platform overview | Gemini Enterprise Agent Platform | Google Cloud Documentation
  47. [47] Gemini Enterprise Agent Platform Tutorial Part 3: Memory & Runtime | Google Cloud - Community
  48. [48] Scale your agents | Gemini Enterprise Agent Platform | Google Cloud Documentation
  49. [49] Set up Memory Bank | Gemini Enterprise Agent Platform | Google Cloud Documentation
  50. [50] AI Agent Memory in 2026: Architectures & Tools Guide | Articles | o-mega
  51. [51] AI Agent Memory 2026: Vector, Graph, Episodic Update
  52. [52] Anthropic adds persistent memory to Claude Managed Agents in public beta | ETIH EdTech News — EdTech Innovation Hub
  53. [53] GitHub - TeleAI-UAGI/Awesome-Agent-Memory: Curated systems, benchmarks, and papers etc. on memory for LLMs/MLLMs --- long-term context, retrieval, and reasoning. · GitHub
  54. [54] Anthropic vs OpenAI 2026: Every Product, Model & Feature Tracked | Value Add VC
  55. [55] Anthropic Release Notes - August 2026 Latest Updates - Releasebot
  56. [56] OpenAI Codex (AI agent)
  57. [57] LoCoMo Benchmark Results 2026 — Dakera AI Memory Engine
  58. [58] AI Memory Benchmarks 2026: LoCoMo, LongMemEval & BEAM
  59. [59] AI Memory Benchmarks 2026: Complete Guide to Evaluating Agent Memory | Bench'd
  60. [60] SuperLocalMemory V3: Information-Geometric Foundations for Zero-LLM Enterprise Agent Memory
  61. [61] Agent Memory Benchmarks 2026: The Real Numbers (LongMemEval, LOCOMO, and the Gaps) | memnode

Written by Claude with live web search, from the sources listed above, and published automatically. Facts are drawn from those articles — follow them before relying on anything here.