OpenAI and AWS shipped their own decision models within two weeks of TypeSafe AI's September 15 launch of Jev, according to a dev.to account. For routing work, the comparison with an LLM call turns on whether a decision model's probabilities are calibrated.
Reality
- Evidence30
- Adoption25
- Hype gap+35
- Incentives55
- Confidence35
Agent runtimes should refuse unit 501 of a 500-unit budget before the call runs, a dev.to post argues. The new AWS and Google Cloud spend caps then become the backstop behind a tighter limit in the agent's own call path.
Reality
- Evidence55
- Adoption20
- Hype gap+15
- Incentives
- Insufficient
- Confidence60
AWS added a per-project spend limit on September 16, 2026 that pauses service at the monthly cap, following Google Cloud's July launch of Spend Caps. A cap that checks each request stops at once, while one built on lagging billing data keeps charging until a function fires.
Reality
- Evidence35
- Adoption
- Insufficient
- Hype gap+10
- Incentives
- Insufficient
- Confidence30
AWS's Bedrock AgentCore Python SDK shipped fixes in v1.6.1 and v1.18.1 for one argument-injection flaw in install_packages(). Teams that let agents or users name packages for Code Interpreter sandboxes should check those names before the SDK sees them.
Reality
- Evidence55
- Adoption
- Insufficient
- Hype gap+10
- Incentives
- Insufficient
- Confidence50
Amazon Quick Sight's new hierarchy filter puts up to five related fields, such as Region down to City, into a single dashboard control. Readers get one control to scan in place of several, and the dashboard author sets the drill order in advance.
Reality
- Evidence60
- Adoption
- Insufficient
- Hype gap+10
- Incentives75
- Confidence65
LEGO-Bench, from the University of Maryland and AWS, scores the best coding agent at 53.4 percent on indoor scenes rebuilt from photos in Blender code. The agents cannot tell when an edit helped, so their revise loop needs a measured score to decide which edits to keep.
Reality
- Evidence45
- Adoption
- Insufficient
- Hype gap−5
- Incentives
- Insufficient
- Confidence40
AWS open-sourced Strands Decider 2B, a model that picks from a fixed list of options in under 150 milliseconds on a local machine, the company says. Agent builders get a small model they can run themselves for routing and tool-selection steps that would otherwise each call a full LLM.
Perspective Coverage
3 publishers
- Builder
- Builder 48%
- Operator
- Operator 25%
- Investor
- Investor 27%
Reality
- Evidence58
- Adoption
- Insufficient
- Hype gap+22
- Incentives60
- Confidence62
Amazon Bedrock bills a Thai customer sentence at 2.4 to 3.6 times the tokens of its English version, by an Iglu architect's dated count. Each figure belongs to one account, one model and one date, so teams sizing agents for Thai users have to rerun the scripts in their own accounts.
Reality
- Evidence55
- Adoption
- Insufficient
- Hype gap+5
- Incentives20
- Confidence50
Amazon will license Synopsys IP and expand its use of Synopsys chip-design tools in a multi-year deal reported at over $1 billion. Synopsys's pledge to tune its simulation software for Trainium and Graviton does more for AWS's position than the license does.
Reality
- Evidence45
- Adoption20
- Hype gap+20
- Incentives75
- Confidence45
Plain Claude Code, with no MCP server or skill, matched AWS's and draw.io's official diagram tools in a five-setup test published on dev.to. Every setup improved as the author kept adding instructions, and the finding covers one architecture on Claude Opus 5, graded by the author.
Reality
- Evidence35
- Adoption
- Insufficient
- Hype gap+15
- Incentives
- Insufficient
- Confidence40
AWS's CloudWatch Omni, generally available since September 23, lets Okta and Entra ID users investigate incidents without AWS console access. CloudWatch dashboard sharing has let outsiders view prebuilt graphs since 2020, so what Omni adds is the investigation itself, one of the reasons teams paid for third-party platforms.
Reality
- Evidence50
- Adoption
- Insufficient
- Hype gap+20
- Incentives60
- Confidence55
Cloudflare says AI-agent requests on its network grew more than 1,700% in a year as non-human traffic passed half the total. It now wants sites to identify and price agents as customers, after advising last year that new domains block AI training crawlers.
Reality
- Evidence35
- Adoption50
- Hype gap+25
- Incentives85
- Confidence40
CTOs told Gergely Orosz that CPU spot pricing, once up to 90% below list, has nearly vanished as AI workloads absorb spare cloud capacity. Teams whose compute budgets assumed spot rates now have to plan for getting capacity at all, on top of paying more for it.
Reality
- Evidence35
- Adoption35
- Hype gap+25
- Incentives30
- Confidence40
AWS's Reimagine 2026 report, built on 154 executive interviews, says review processes sized for six-month IT programs are driving AI use underground. AWS's answer is governance built into the systems themselves, with humans kept accountable for outcomes.
Reality
- Evidence40
- Adoption25
- Hype gap+10
- Incentives65
- Confidence45
EKS Pod Identity, launched at re:Invent 2023, replaces IRSA's per-cluster OIDC trust with one pods.eks.amazonaws.com principal on every IAM role. A dev.to migration guide has each role trust both paths during cutover, so teams can move and verify one workload at a time.
Reality
- Evidence45
- Adoption
- Insufficient
- Hype gap+15
- Incentives
- Insufficient
- Confidence50
FastGPU's September 27 snapshot of 28 GPU clouds puts the cheapest hyperscaler H100 at $5.38 an hour, 3.0x the $1.79 market floor. That premium buys IAM, managed services and credits, and it is worth paying when a team actually uses them.
Reality
- Evidence45
- Adoption
- Insufficient
- Hype gap+15
- Incentives65
- Confidence50
An interview with Lean's founder says an AI translated C zlib into Lean and proved compress-then-decompress round-trips in about a week. The remaining work is optimization without breaking the proof.
Reality
- Evidence25
- Adoption35
- Hype gap+40
- Incentives65
- Confidence30
Throughline's on-call agent returns a three-way coverage verdict with every memory recall, so a timed-out search cannot pass as "no prior incidents". Its author hit the same error-as-empty trap in CockroachDB's managed MCP server, where failures come back as HTTP 200.
Reality
- Evidence35
- Adoption
- Insufficient
- Hype gap−10
- Incentives30
- Confidence40
AWS's Java extended clients for large SQS and SNS messages pull Jackson versions with seven known advisories via a library last released in March 2024. Kotlin services now have ports on Maven Central that take Jackson out of that code path.
Reality
- Evidence45
- Adoption
- Insufficient
- Hype gap+20
- Incentives70
- Confidence45
JFrog has documented an on-behalf-of exchange so agent calls reach Artifactory as the signed-in user. The price is the Gateway's cached tool search, which per-user discovery gives up.
Reality
- Evidence55
- Adoption
- Insufficient
- Hype gap+15
- Incentives70
- Confidence60
Earlier coverage
- AWS puts the agent runtime behind an n8n node, and leaves the permissions with you
Build · August 24, 2026 · 1 publisher
- Darktrace catches an AI agent hacking its own grader to fake a perfect score
Invest · September 25, 2026 · 1 publisher
- HyperPod holds the inference deployment until every target node has cached the weights
Build · September 10, 2026 · 1 publisher
- Hybrid ML-KEM key exchange adds about 1,600 bytes to every fresh TLS handshake
Build · September 14, 2026 · 1 publisher
- AWS says the damage in Bahrain outran what multi-AZ was designed to absorb
Build · September 20, 2026 · 2 publishers
- Multi-cloud portability breaks on how each provider answers the same call
Product · September 25, 2026 · 1 publisher
- Copilot's hard stop on AI coding spend ships switched off
Build · September 24, 2026 · 1 publisher
- Docker hands CNCF a spec that makes an agent's permission list part of the image
Product · September 24, 2026 · 1 publisher
- Choosing US-only inference for Kimi K3 costs about 11% more than Bedrock's global profile
Build · September 23, 2026 · 1 publisher
- AWS claims its pre-assembled agent harness cuts costs 28% on the same Claude and GPT models
Product · September 23, 2026 · 1 publisher
- Amazon moves seller pricing and inventory controls into Anthropic's Claude
Product · September 23, 2026 · 1 publisher
- ShinyHunters claims a PeopleSoft zero-day gave it code execution on FBI servers
Product · September 23, 2026 · 1 publisher
- Amazon courts the workers it laid off for its AI and cloud openings
Leadership · September 23, 2026 · 1 publisher
- Luna lands at one tenth of Terra's price on both input and output tokens
Build · September 22, 2026 · 1 publisher
- A departed employee's GitHub OAuth token cloned about 170 CrowdSec repositories in nine minutes
Build · September 22, 2026 · 2 publishers
- A CreateAIBenchmarkJob call ramps a SageMaker endpoint from 64 to 1,024 concurrent requests
Build · September 22, 2026 · 1 publisher
- Okta puts an enforcement point between AI agents and the tools they call
Product · September 22, 2026 · 1 publisher
- Tagging tool calls with source at write time turns the reverse audit into one traversal
Build · September 21, 2026 · 1 publisher
- Strands Harness keeps five subsystems local and routes one call to Bedrock
Build · September 21, 2026 · 1 publisher
- Writing the edges in Python takes the routing decision out of the LLM call
Build · September 21, 2026 · 1 publisher
- A Lambda transform lands blocked prompt injections in the same Athena catalog as CloudTrail
Build · September 21, 2026 · 1 publisher
- AWS automatically quarantines leaked IAM keys and opens a support case for the affected user
Security · September 21, 2026 · 1 publisher
- Mystery model Union Alpha hit a billion tokens a minute before vanishing from listings and being revealed as Pareto
Build · September 20, 2026 · 1 publisher
- Anthropic pledges more than $100bn to AWS for $5bn of Amazon cash now
Invest · September 19, 2026 · 1 publisher
- Kimi K3 puts explicit prompt caching on Bedrock behind a 1,024-token minimum prefix
Build · September 18, 2026 · 1 publisher
- Amazon moved a subset of its servers back to a five-year depreciation life
Build · September 18, 2026 · 1 publisher
- Kiro and Claude Code both picked a TGI container that could not load Qwen3
Build · September 18, 2026 · 1 publisher
- Saudi Arabia trims NeoCity and LIV golf while pledging $15 billion to domestic AI
Leadership · September 18, 2026 · 1 publisher
- Anthropic buys a decade of Trainium capacity with a $100 billion commitment to AWS
Leadership · September 18, 2026 · 1 publisher
- HyperPod's inference gateway scores KV cache and LoRA residency before it picks a pod
Build · September 18, 2026 · 1 publisher
- AWS's FinOps Agent reads CloudTrail to name the role behind the cost spike
Build · September 17, 2026 · 1 publisher
- AWS puts Gemma 4 on Bedrock inside its EU-only Brandenburg region
Build · September 17, 2026 · 1 publisher
- AWS's new sign-up flow hides two of the three accounts it creates
Security · September 17, 2026 · 1 publisher
- AWS's Git metrics pipeline baselines AI coding tools on commit and pull request counts
Build · September 17, 2026 · 1 publisher
- AWS measures its 38 healthcare agent skills against the same agents without them
Build · September 16, 2026 · 1 publisher
- Apple weighs building its M8 Ultra servers around Nvidia's NVLink Fusion interconnect
Leadership · September 16, 2026 · 1 publisher
- Blocking checkpoint writes took up to 40 percent of wall time on AWS's H100 test runs
Build · September 16, 2026 · 1 publisher
- One blueprint instruction separates the patient's date of birth from the signature date
Build · September 16, 2026 · 1 publisher
- AWS gives its system prompt optimizer a shell over a directory of scored traces
Build · September 16, 2026 · 1 publisher
- A single Iceberg tag switches off S3 Tables snapshot expiry for the whole table
Build · September 16, 2026 · 1 publisher