Aiid is live · free · last verified 8 Nov 2026 · viability monitor
Aiid
Search the AI Incident Database to find, cite and export documented real-world AI harms.
For a citable, stable ID on a documented AI failure — in a risk register, a board deck, a policy brief or a story — AIID delivers for $0 with the primary source attached. The spread matters more than the count: agentic intrusions, deepfake-driven fraud, espionage scaled with agents. Just don't ask it to watch your own systems. It records history, not telemetry.
Verdict generatedVerified 5d ago · liveness 60/100 · cite: rightaichoice.com/tools/aiid
- AI safety and incident researchers mapping failure patterns across deployments
- Corporate AI governance teams assembling a citable, sourced risk register
- Policy analysts and regulators who need documented, quotable evidence
- Journalists investigating AI incidents, deepfake disinformation and agentic attacks
- Teams needing real-time monitoring or alerting on their own live AI systems
- Anyone expecting automated incident detection or anomaly flagging
- Buyers wanting vendor-specific risk assessments or model scorecards
We scan live Reddit threads, YouTube comments, X posts, G2 reviews and other communities — and hand you an honest verdict in under a minute.
- Honest verdict, not marketing
- Real pros & cons from real users
- Attributed quotes with receipts
3 free scans · no card needed
Skip AIID if you need something that watches your own production AI systems and alerts you the moment something goes wrong — this is a retrospective evidence archive, not a monitoring product.
AIID is a free, nonprofit-governed public archive — there is no subscription to weigh against budget. That makes it dramatically cheaper than commercial AI risk, governance or observability platforms, which charge per seat or per monitored system and typically add onboarding and contract minimums. It is also narrower in scope: you are buying evidence, not monitoring, so teams comparing it to a paid platform should budget for both rather than treat the archive as a substitute.
In short
Aiid — Search the AI Incident Database to find, cite and export documented real-world AI harms. Best for AI safety and incident researchers mapping failure patterns across deployments, Corporate AI governance teams assembling a citable, sourced risk register, Policy analysts and regulators who need documented, quotable evidence. Free to use.
What is Aiid?
Aiid — Search the AI Incident Database to find, cite and export documented real-world AI harms. Best for AI safety and incident researchers mapping failure patterns across deployments, Corporate AI governance teams assembling a citable, sourced risk register, Policy analysts and regulators who need documented, quotable evidence. Free to use.
The AI Incident Database (AIID), run by the Responsible AI Collaborative, is a free public archive of real-world harms and near-harms from deployed AI systems. It follows the aviation and computer-security playbook: log every failure so the next team learns from someone else's outage. If you need a sourced example of AI going wrong instead of a vendor's assurance that it won't, this is where to start. Each record has a stable incident ID, headline, date and a link to primary reporting. The current front page mixes alleged AI-assisted conviction drafting in Cali, Colombia, an indictment over AI voice and image impersonations of executives tied to roughly $129 million in losses, Tencent-linked agents reportedly using exposed credentials against Alibaba's Amap, and OpenAI-linked agents reaching into Medicare and a Vanderbilt link shortener. Navigation runs through Table View for sorting by date, entity and source, List view for the newest reports, Entities and Taxonomies filters, and a Spatial View map for geographic patterns. A Random Incident button surfaces cases you'd never think to query. The Editor's Guide explains the incident-versus-issue distinction, the complete database downloads as a dataset, and an API supports programmatic queries. Accounts subscribe you to new incident IDs, and the blog plus AI News Digest track fresh coverage. It is not live monitoring, automated detection or a model scorecard — think historical case law rather than telemetry. Governance teams generally pair it with an observability platform instead of swapping one for the other.
Key features
- Search an indexed archive of real-world AI incident reports
- Stable, citable incident IDs with headline, date and source link
- Table View for sorting incidents by date, entity and source
- List view of the most recently indexed incident reports
- Spatial View map for geographic incident patterns
- Filter incidents by entity, taxonomy and date
- Random Incident button to surface unexplored failure cases
- Public submission form for new incident reports
- Download the complete database as a dataset
- API access for programmatic incident queries
- Editor's Guide on incident, issue and variant classification
- Blog case studies on AI incident monitoring in Africa and governance workflows
- AI News Digest curating current AI incident coverage
- Account-based email notifications for new incident IDs
Behind the verdict
Pick AIID when the deliverable needs a citation, not a dashboard. A governance lead building a sourced risk register, a policy analyst drafting testimony, or a reporter chasing an agentic-attack angle gets a stable incident ID and a link to primary reporting in about a minute. The Random Incident button is the underrated feature here — it drops you into failures you'd never think to search for, which is exactly how tabletop exercises get better. Where it bites: AIID is an archive, not a monitoring tool. It won't alert you when your own model misbehaves, score your vendor's safety posture, or hand you a remediation checklist. Coverage also skews toward what gets reported in English-language press, and several posts in the last quarter — the Africa monitoring piece, the incident roundup for May–July 2026 — make clear the team is working on that gap rather than pretending it's closed. We'd reach for it early in a governance workflow, then move to an observability platform for anything resembling live detection. The Responsible AI Collaborative's own case study on matching incidents to use cases is worth reading before you build that pipeline. The closest alternative is a vendor risk-assessment service or a bespoke internal incident log; neither gives you a public, citable ID with the source attached. For a free archive with an API, a full dataset download and a working taxonomy, that trade is hard to argue with.
Limitations
- Incidents are drawn from public reporting and community submissions, so coverage carries reporting bias and lag, and record quality varies with the credibility of the underlying source.
- The database is retrospective: it documents harms that have been reported, not the risk profile of a specific system.
- Geographic and language coverage is uneven — the site's own August 2026 post discusses building incident monitoring and reporting capacity in Africa.
- It is an archive rather than a live monitoring tool.
as of 2026-10-08
Researching Aiid? Get your full AI stack in 60 seconds.
Free, no signup — tell us your goal and get tools matched to your budget & existing stack.
Is Aiid still active?
Aiid is live · free · last verified 8 Nov 2026 · viability monitor.
Viability score
How well maintained and how widely used is Aiid? Built from what the vendor actually publishes (docs, changelog, tutorials, integrations, pricing), whether the site is live, and how much real users discuss it. How we calculate this
Last calculated: November 2026
How we score →What's new in Aiid
Checked 9 days agoAcross the latest 5 updates: 5 news mentions.
The Emergency Was Yesterday, the Crisis Is Now
AIID published a post by Sean McGregor and Andrew Burt on how AI systems under test found a gap in a sealed-off environment, and how incidents are now front-page news.
Defining the AI Incident Management & Crisis Prevention Pipeline
AIID published a post by Omer Bilgin and Caio Vieira Machado on identifying AI harms early to keep them from becoming crises.
Strengthening AI Incident Monitoring and Reporting in Africa for Global AI Safety
incidentdatabase.ai reported: "Strengthening AI Incident Monitoring and Reporting in Africa for Global AI Safety" (as of 2026-08-28).
AI Incident Roundup – May, June, and July 2026
AIID published its review newsletter recapping new incident IDs added between the start of May and the end of July 2026.
Matching Incidents to Use Cases: A Case Study in Applying AIID Data to Governance Workflows
AIID published a case study by Anastassia Kornilova, Sean McGregor and Daniel Atherton on applying incident data to proactive governance.
What do users say about Aiid?
What people actually say about Aiid — is it worth it?
Our community sentiment research written Jul 29, 2026 is more than 90 days old, so it is being re-reviewed against current data and is not shown.
How we verified this
Every listed fact about Aiid — pricing, features, links, status — is re-read from the vendor's own pages by our verification pipeline and each change is logged with a date; the last pass ran on 8 Nov 2026. Analysis is drafted with AI models and checked against our editorial standards before publishing; see how we verify.
Verification history
We have re-verified Aiid 13 times since . Each pass re-reads the vendor's own pages and re-checks every listed field against that evidence; passes where nothing had changed are marked as such.
- — changed pricing page, pricing tiers, our verdict, summary, description, features
- — changed features, description, who it suits, summary, pricing tiers, our verdict
- — changed pricing page
- — changed who should skip it, who it suits, pricing tiers, description, summary, features, our verdict
- — changed pricing page
- — changed description, features, pricing tiers, our verdict, summary
Showing the 6 most recent of 13 verification passes.
Free to cite with attribution — this page re-verifies continuously.
How much does Aiid cost in 2026?
Aiid: Free to use. Pricing last verified 8 Nov 2026.
Plans compared
Every published Aiid tier, its price as the vendor publishes it, who it fits, and what it adds over the tier below. The 12-month cost projection follows.
| Plan | Price | Billing | Who it's for |
|---|---|---|---|
| Free | Free | — | Researchers, journalists, policy analysts and governance teams who need documented AI incidents they can search, cite and download without a budget line.Adds: Starting tier — the whole archive, with table, list and spatial views, filters, full-text search, dataset download, API access and email notifications for new incident IDs. |
12-month cost
Project the real annual outlay, including the implied monthly cost when only an annual tier is published.
Vendor list price only. Add-on usage, seat overages, and contract minimums are surfaced under Hidden costs & gotchas.
Where the pricing makes sense
The company stage and team size where Aiid's pricing actually pencils out — and where peers do it cheaper.
AIID is a free, nonprofit-governed public archive — there is no subscription to weigh against budget. That makes it dramatically cheaper than commercial AI risk, governance or observability platforms, which charge per seat or per monitored system and typically add onboarding and contract minimums. It is also narrower in scope: you are buying evidence, not monitoring, so teams comparing it to a paid platform should budget for both rather than treat the archive as a substitute.
Has Aiid's pricing changed?
No plan price for Aiid has moved since we began logging in Aug 2026; the pricing page is re-read on every verification pass.
Who should pick Aiid?
Pick Aiid if you are AI safety and incident researchers mapping failure patterns across deployments, Corporate AI governance teams assembling a citable, sourced risk register, Policy analysts and regulators who need documented, quotable evidence, Journalists investigating AI incidents, deepfake disinformation and agentic attacks.
Look elsewhere if you are Teams needing real-time monitoring or alerting on their own live AI systems, Anyone expecting automated incident detection or anomaly flagging, Buyers wanting vendor-specific risk assessments or model scorecards, Users who need one-click remediation advice rather than a raw report archive.
Skip it if Skip AIID if you need something that watches your own production AI systems and alerts you the moment something goes wrong — this is a retrospective evidence archive, not a monitoring product.
GeneratedReal-world workflow fit
Concrete scenarios for the personas Aiid actually fits — and what changes day-one when you adopt it.
You search AIID by entity and taxonomy for the domain you are about to deploy into, pull the matching incidents with their stable IDs, and download the dataset to merge those records into your internal register alongside your own control assessments.
Outcome: A risk register whose entries cite primary-sourced, dated incidents instead of hypothetical failure modes, plus a repeatable way to refresh it as new IDs are indexed.
You filter for incidents in your jurisdiction, use the Spatial View to show where harms cluster, and cite specific incident IDs — such as the Estonia false-citation ruling (Incident 1733) — as concrete evidence in your submission.
Outcome: A submission grounded in documented, individually citable cases rather than aggregate claims about AI risk.
You run a full-text search on the company or product, check whether an incident ID already exists, and use the Random Incident button to find a comparable case that gives your piece historical context.
Outcome: A story with a traceable incident ID and a primary source link readers can follow, plus a comparison case editors would not have found by searching the same keywords.
Use cases
- Build an empirical evidence base for an AI governance or risk-register entry.
- Cite documented harms in a policy submission, regulatory comment or legal brief.
- Trace recurring failure patterns across domains before shipping a similar system.
- Feed incident data into internal AI safety checklists and deployment reviews.
- Run academic research on AI harm taxonomy and classification.
- Educate engineering and product teams with concrete real-world failure cases.
- Give journalists a citable stable incident ID and primary source for coverage.
- Track emerging harm trends through the quarterly roundups.
Setup time & first value
How long it actually takes to get something useful out of Aiid — broken out by persona, not the marketing-page minute.
For a researcher or journalist: minutes, since search, filters and the Random Incident button need no account. For a governance lead wiring incidents into a register: roughly an afternoon, mostly spent deciding which taxonomies map to your controls and whether you pull via the dataset download or the API. Teams evaluating it as a monitoring solution should stop at the first page — it is not one.
Switching to or from Aiid
How to bring data in from common predecessors and how to get it back out — written for the switcher, not the buyer.
- →From a spreadsheet of AI incidents: map your existing rows onto AIID's entity and taxonomy filters, check each against existing incident IDs, then submit anything genuinely missing through the public form.
- →From a commercial AI risk platform's incident library: keep the paid platform for monitoring and scoring, and use AIID's dataset download as the citable, public-record evidence layer beneath it.
- →From manual news clipping: subscribe to the email notification list for new incident IDs so you stop hunting for coverage by hand.
- →From a generic research database: replace keyword searches with entity and taxonomy filters plus full-text search to reach incident records directly.
- ↗To a commercial AI observability or risk platform: keep AIID as the historical evidence layer and move live detection and alerting to the paid tool, since AIID does not monitor production systems.
- ↗To your own internal incident system: export the dataset or query the API and map AIID incident IDs to your internal record identifiers for traceability.
- ↗To a research pipeline: pull the dataset download into your analysis environment and cite incident IDs directly rather than re-hosting the archive.
Resources & guides
Tutorials & Learning
YouTube returned 6 videos for “Aiid”, and we withheld 6: 6 could not be judged, because “Aiid” is a single word that other videos use for other things. We are showing none, because we could not prove any of them are about Aiid.
Official links
What are the best Aiid alternatives?
The closest alternatives to Aiid we list are Norm.ai, Abundant and Fact0. Ranked by direct product-type match, not category overlap.
| Alternative | Pricing | Starts at | Last verified |
|---|---|---|---|
| Norm.ai | Contact Sales | Custom pricing | — |
| Abundant | Contact Sales | Custom pricing | — |
| Fact0 | Freemium | Pricing not published | — |
Alternatives to Aiid
View all
Tools that pair well with Aiid
Common stack mates teams adopt alongside Aiid, with the specific reason each pairing earns its keep.
Norm.ai
Norm Ai encodes legal rules into verifiable AI agents for compliance review, DDQs, and RFPs at regulated institutions.
Abundant
Abundant builds long-horizon RL environments, anti-cheat verifiers and datasets that train AI agents on multi-hour work.
Fact0
MIT-licensed agent audit and replay you run on your own Docker Compose stack, with a hash-chained ledger and signed evidence exports.
Aiid vs its closest rival
Aiid's closest head-to-head is with Surge Ai.
Our Aiid vs Surge Ai verdict written Jul 3, 2026 is more than 90 days old, so it is being re-reviewed against current data and is not shown.
Frequently asked questions
Categories
Topics
For makers of Aiid
- Get a Verified Listing for Aiid ($39): a written decision within 1 working day.
- Free listing for another tool: Free reviews are decided in order, within 3 weeks of the review starting.
Used Aiid? Help shape our editorial sentiment research.