SOC 2 access reviews credit humans with actions their AI agents took
SOC 2 controls CC6.1 to CC6.3 let AI agents act under human credentials without failing a check, a BleepingComputer column argues. A clean review can leave responders unable to say who ran a query, and the criteria already let firms register agents as their own identity.
Reality
- Evidence30
- Adoption
- Not scored
- Hype gap+35
- Incentives80
- Confidence35
buildOne report1 publisher A recovery-decision event can be written before the provider's own event id exists
An identity provider's history proves a factor challenge succeeded, so SOC 2 evidence for what the application decided next has to come from its own append-only event, with the policy version recorded at the moment of the decision.
Reality
- Evidence45
- Adoption
- Not scored
- Hype gap+10
- Incentives20
- Confidence50
The startup's agents draft security policies and gather audit evidence, and its founders say a person still approves each policy while an independent auditor still signs the report.
Perspective Coverage
3 publishers
- Builder
- Builder 35%
- Operator
- Operator 35%
- Investor
- Investor 30%
Reality
- Evidence45
- Adoption52
- Hype gap+18
- Incentives74
- Confidence58
buildOne report1 publisher Astronomer's 84 percent planning figure counts only the respondents still on Airflow 2
Astronomer's State of Apache Airflow 2026 puts 5,818 practitioners behind two migration numbers measured on different denominators. Recomputed on one base, about 62 percent are planning a move they have not started.
Reality
- Evidence46
- Adoption55
- Hype gap+30
- Incentives78
- Confidence56
buildOne report1 publisher A resource-by-resource AWS audit passes an account with GuardDuty disabled
In HackerOne report #3022516 the trail was multi-region, the metric filter for unauthorised API calls existed and the alarm published to an SNS topic, while threat_detection.enabled read false. Turning it on costs one CLI call.
Reality
- Evidence55
- Adoption
- Not scored
- Hype gap+10
- Incentives65
- Confidence60
buildOne report1 publisher An unscoped GitHub token and an unprotected branch produce the same SOC 2 finding
Suman Shashikumar's open-source evidence collector reads AWS and GitHub controls and labels each finding with the criterion it proves. The branch-protection check is where the labelling gets hard.
Reality
- Evidence46
- Adoption6
- Hype gap−8
- Incentives55
- Confidence52
Wyoming details pattern of LayerZero failures, including a mismanaged key, behind its exit to Chainlink's CCIP
The Stable Token Commission's September 14 memo opens with a $292 million North Korea-linked bridge theft and ends with two access-control lapses under Wyoming's own token, 149 days after the attack that started the review.
Reality
- Evidence38
- Adoption55
- Hype gap+22
- Incentives74
- Confidence44
buildOne report1 publisher One add_header in an nginx location block discards the other four set above it
A dev.to comparison of hand-written nginx directives against delegated edge policy comes down to the inheritance and duplication rules that make a passing config file a poor description of what a client actually receives.
Reality
- Evidence58
- Adoption
- Not scored
- Hype gap−5
- Incentives20
- Confidence50
Amazon's cure for shadow AI ships on the desktop from outside its European sovereign cloud
AWS has made the Quick desktop assistant generally available on macOS and Windows and pitched it to IT as the sanctioned answer to shadow AI. It runs in Frankfurt, London and Ireland, and not in the Brandenburg sovereign cloud.
Reality
- Evidence55
- Adoption34
- Hype gap+28
- Incentives74
- Confidence55
Polymath names a 177-day-old trust company as custodian for its tokenized securities
High Ridge Trust, a Nevada-based trust company open since March, will hold securities issued on Polymath's Polymesh chain. The partnership was struck 25 days into TruGolf's all-share takeover of Polymath.
Reality
- Evidence26
- Adoption
- Not scored
- Hype gap+46
- Incentives74
- Confidence41
buildOne report1 publisher Amazon pitches the Quick desktop app to IT as containment for shadow AI
The argument AWS makes to IT for its newly general desktop assistant is that unapproved tools already move data out of controlled environments, and its answer is audit trails in CloudWatch and CloudTrail plus four named certifications.
Reality
- Evidence26
- Adoption24
- Hype gap+46
- Incentives93
- Confidence62
buildOne report1 publisher Context7 turns its retrieval log into a documentation hosting business
Docs7 takes a Mintlify config, serves Markdown from the URLs agents already request, and counts which AI products fetch what, a metric Context7 itself says measures provider visits rather than readers.
Reality
- Evidence36
- Adoption18
- Hype gap+28
- Incentives80
- Confidence44
buildOne report1 publisher CloudFormation delegates ASG instance replacement to the Instance Refresh engine
A native AutoScalingInstanceRefresh update policy makes the stack the single source of truth for AMI rollouts, which also puts Instance Refresh's abort-and-rollback behaviour on the critical path of every stack update.
Reality
- Evidence30
- Adoption15
- Hype gap+10
- Incentives25
- Confidence32
buildOne report1 publisher Attacker asked a researcher's personal 'vibe-coded' agent dashboard for the model provider's API key
An AI evaluator with frontier-model access lost a provider key from an experiment on a researcher's personal EC2 box. The reach of the credential, not the seriousness of the project, decided what the mistake cost.
Reality
- Evidence60
- Adoption58
- Hype gap+12
- Incentives66
- Confidence62
buildOne report1 publisher Vanta's agentic controls slot into the ISO program you already run
The library maps to ISO 27001 and 42001 rather than asking for a new certificate, which is why trying it costs a scoping meeting. The 22 buyer-side controls are the half that will end up in vendor questionnaires.
Reality
- Evidence55
- Adoption22
- Hype gap+15
- Incentives74
- Confidence45
buildOne report1 publisher Team-scoped gateway keys tag org, team and user identity for AI spend dashboard
A dev.to walkthrough replaces developer-held provider keys with a gateway that tags every call with org, team and user. The tagging is server-side, which is the right call, and the dashboards behind it refresh every 15 minutes.
Reality
- Evidence23
- Adoption
- Not scored
- Hype gap+34
- Incentives91
- Confidence52
buildOne report1 publisher Same-day GPT-5.6 on Azure kills the parity argument, leaving auth and residency to decide
Azure used to trail new OpenAI models by four to eight weeks. With the GPT-5.6 family shipping on both platforms at once, the remaining differences are credentials, regions and about 15 to 40 percent of overhead.
Reality
- Evidence24
- Adoption
- Not scored
- Hype gap+34
- Incentives38
- Confidence31
Parallels ships OpenGL 4.3 on Apple silicon, and the fleet question moves to which chip you own
Parallels Desktop 27 puts a Metal-based driver behind Windows VMs, but the numbers it quotes are scoped to M3 and M4 Macs. The compatibility win is real; the throughput win is conditional.
Reality
- Evidence30
- Adoption22
- Hype gap+38
- Incentives72
- Confidence34
150 Jenkins masters, one control plane: the fix for CI sprawl was not a migration
A platform team put a dashboard, idempotent Ansible and Prometheus over 150+ Jenkins instances instead of replatforming. The bill was 20 months, and the savings are still unmeasured.
Reality
- Evidence28
- Adoption24
- Hype gap+34
- Incentives68
- Confidence33
Enterprise security reviews went from 20 questions to hundreds of rows, and vendors pay first
One vendor reports buyer questionnaires growing at least tenfold, now demanding customer-controlled identity, exportable logs, data residency and a no-AI-training promise. The build comes before the signature.
Reality
- Evidence22
- Adoption15
- Hype gap+32
- Incentives82
- Confidence34