Whitepapers

The AI Testing Newsletter,
read in full.

Every issue, as a full article. What AI vulnerability discovery is doing to product security, with the numbers and incidents that prove it. Also on LinkedIn, new issues monthly.

Proof Over Probability

Proof Over Probability, weekly.

One issue a week on AI, exploitability, and FDA cybersecurity review. No spam, unsubscribe anytime.

Issue 12 · Agentic remediation

The finding loop outran the fixing loop

An autonomous agent took the number one spot on HackerOne's US bug bounty leaderboard in June 2025, with around 1,060 valid submissions. In December 2025 a research agent named...

2026-08-20 · 4 min read
Issue 11 · Threat modeling

A graph is what every threat model wants to be

Most triage arguments I sit through are really arguments about whether an edge exists. Can an attacker reach the vulnerable function? Does anything authenticate that path? Does...

2026-08-11 · 3 min read
Issue 10 · Agent containment

An AI agent broke out of its sandbox and hacked a company

This one is worth stopping on. On July 21 OpenAI disclosed that two of its own models, GPT-5.6 Sol and an unnamed, more capable pre-release model, escaped an isolated test...

2026-07-21 · 4 min read
Issue 9 · FDA human factors vs cyber gap

FDA's cyber guidance keeps pointing at human factors. Human factors never points back.

FDA finalized 'Content of Human Factors Information in Medical Device Marketing Submissions' on May 29, 2026, replacing the December 2022 draft, with a town hall set for July...

The AI Testing Newsletter · 4 min read
Issue 8 · AI guardrails as a blocklist

Guarding the guardrails: AI reached for a 2005 idea

On July 1, 2026 Anthropic redeployed Fable 5 globally, with Mythos 5 restored to its Project Glasswing partners, after Commerce lifted the export ban; Fable had launched June 9...

The AI Testing Newsletter · 6 min read
Issue 7 · AI model choice as supply-chain risk

Cybersecurity Testing AI Model Arms Race Heats Up

The Wall Street Journal covered the Fable 5 and Mythos story from the model arms-race angle, framing the Commerce Department's move as partly driven by concern that a...

The AI Testing Newsletter · 3 min read
Issue 6 · AI export ban and dual-use defense

Technical breakdown of the export ban on cybersecurity testing

On June 13, 2026 the US Commerce Department issued an export control directive suspending Fable 5 and Mythos 5 for every foreign national, including foreign national employees...

The AI Testing Newsletter · 2 min read
Issue 5 · FDA AI-enabled device cyber testing

Your AI-enabled medical device is about to get a deficiency for cyber testing

In 2025 FDA cleared 295 AI/ML-enabled devices, roughly 97% of them through 510(k), and most almost certainly did not perform the cybersecurity testing FDA's January 7, 2025...

2026-06-26 · 5 min read
Issue 4 · Export control and model dependency

The day the frontier went dark

At 5:21pm ET the night before publication, Anthropic received a US government export control directive and within hours disabled Fable 5 and Mythos 5 for every customer: not...

2026-06-13 · 4 min read
Issue 3 · Agentic AI cost and quality variance

The AI Casino: Why Agentic Pipelines for Product Security Are a Gamble You Can Lose in Minutes

Running agentic AI pipelines on frontier models behaves like a casino: a paid B2B service where quality is neither guaranteed nor measurable, the operator can change the game...

2026-05-30 · 8 min read
Issue 2 · AI vuln discovery harness architecture

Chatbot or Vulnerability Discovery Tool?

The model is not the tool; the harness is the tool. The model provides reasoning and the harness turns reasoning into work against a real target, deciding which tools the model...

The AI Testing Newsletter · 9 min read
Issue 1 · AI vuln discovery in medtech

Myth about Mythos

The legacy triage model (a CVE drops, spend two days deciding if it applies, log a row in the cyber risk spreadsheet, repeat ~15-20 times a month) is finished now, not in five...

The AI Testing Newsletter · 8 min read
Issue 9 · FDA human factors vs cyber gap

FDA's cyber guidance keeps pointing at human factors. Human factors never points back.

FDA finalized 'Content of Human Factors Information in Medical Device Marketing Submissions' on May 29, 2026, replacing the December 2022 draft, with a town hall set for July...

Newsletter brief · PDF download
Follow along

The newsletter keeps publishing.

New issues are condensed into whitepapers here as they ship. For the full essays as they land, follow ELTON Cyber on LinkedIn.

Get started

Reading is good. Evidence is better.

See how the platform behind the newsletter verifies exploitability on real devices, with every disposition evidenced.

Exploitability management for medical devices. FDA §524B methodologyExploitability proven at runtime95% faster than legacy testing Book a Demo
Platform
Platform OverviewDigital TwinAutonomous TestingExploitability VerificationFind the 1%Remediation OptimizationELTON TestLink™SBOM, VEX & ReportingCVSSv4 MigrationProduct Tour
Solutions
Postmarket SurveillanceIncident ResponseSecurity EngineeringRegulatory AffairsFDA §524BEU MDR/CRAEU REDNIS2IMDRF N60 / N73Japan MHLW
Why ELTON
Why ELTONProof Over ProbabilityFind the 1%AI PentestingMDDT MethodologyCredentialsDevice ModalitiesPricingELTON vs. Legacy Testing
Resources
FDA Deficiency ListFDA Testing RequirementsFDA Cyber SOPs & TemplatesRemediation LibraryRegulatory GuidesWebinarsThe End of Legacy TestingThe AI Vulnerability ExplosionSecurity AdvisoriesWhitepapersIntelligence & Blog
Company
AboutLeadershipCareersContact Book a Demo