AI can shorten the path from a code change to a running program, but a trusted author, signed artifact, or clean component list does not prove what that program will do when it runs. Ken Ammon’s proposal is to extend Zero Trust thinking to code: check whether its behavior is allowed before execution, while keeping existing review, scanning, provenance, and access controls in place.
What is AI malware?
“AI malware” can refer to malicious software created or adapted with AI, but that label alone does not establish how the software was made, how widely it is used, or whether AI materially changed an attack. NIST’s March 2025 report, Adversarial Machine Learning: A Taxonomy and Terminology of Attacks and Mitigations, defines terminology for attacks on machine-learning systems, including methods, life-cycle stages, attacker goals, capabilities, and mitigations. It is a taxonomy, not a measurement of AI malware prevalence.
As an Amazon Associate I earn from qualifying purchases.
Ammon’s argument is less about proving a wave of AI-authored malware than about the security consequences of faster code generation and change. He contends that AI can compress the time available for review, scanning, testing, approval, and deployment, and can make payloads easier to vary. Those are the author’s risk claims, not a quantified demonstration that attackers universally move faster than defenders.
How does AI-generated code change cybersecurity?
Code that is generated, modified, or invoked quickly can pass through familiar safeguards before people have time to understand its behavior. The concern applies not only to a finished application, but also to scripts, dependencies, editor extensions, build workflows, and code produced or acted on by an agent. More code and faster changes increase the importance of knowing what is allowed to execute and with which permissions.
#1 Best Overall
- POWERFUL SECURITY KEY: The Security Key C NFC is the essential physical passkey for protecting your digital life from phishing attacks. It ensures only you can access your accounts.
- WORKS WITH 1000+ ACCOUNTS: Compatible with Google, Microsoft, and Apple. A single Security Key C NFC secures 100 of your favorite accounts, including email, password managers, and more.
- FAST & CONVENIENT LOGIN: Plug in your Security Key C NFC via USB-C and tap it, or tap it against your phone (NFC) to authenticate. No batteries, no internet connection, and no extra fees required.
- TRUSTED PASSKEY TECHNOLOGY: Uses the latest passkey standards (FIDO2/WebAuthn & FIDO U2F) but does not support One-Time Passwords. For complex needs, check out the YubiKey 5 Series.
- BUILT TO LAST: Made from tough, waterproof, and crush-resistant materials. Manufactured in Sweden and programmed in the USA with the highest security standards.
That does not make conventional controls obsolete. Code review, testing, scanning, signatures, provenance, and dependency inventories answer important questions about an artifact or its development history. Ammon’s point is that those checks do not, on their own, establish that every action the code takes at runtime is acceptable.
Can Zero Trust be applied to code?
Yes, as a design approach: treat execution as a decision that needs evidence and policy, rather than assuming that code is safe because it came from a trusted place or passed a previous check. In this proposal, a system evaluates whether a program’s expected behavior and requested privileges are permitted before allowing it to run, then monitors execution.
Rank #2
- POWERFUL SECURITY KEY: The YubiKey 5C NFC is the most versatile physical passkey, protecting your digital life from phishing attacks. It ensures only you can access your accounts
- WORKS WITH 1000+ ACCOUNTS: Compatible with popular accounts like Google, Microsoft, and Apple. A single YubiKey 5C NFC secures 100+ of your favorite accounts, including email, password managers, and more
- FAST & CONVENIENT LOGIN: Plug in your YubiKey 5C NFC via USB and tap it, or tap it against your phone (NFC), to authenticate. No batteries, no internet connection, and no extra fees required
- MOST SECURE PASSKEY: Supports FIDO2/WebAuthn, FIDO U2F, Yubico OTP, OATH-TOTP/HOTP, Smart card (PIV), and OpenPGP. That means it’s versatile, working almost anywhere you need it
- PRIMARY & SPARE KEYS: Just like having a spare house key, we recommend buying two YubiKeys - one for daily use and one as a spare. That way you’ll never get locked out of your accounts
This is an extension of Zero Trust principles, not a recognized universal “Zero Trust for Code” standard. NIST’s SP 1800-35, published in June 2025, is an implementation guide for enterprise Zero Trust Architecture consistent with SP 800-207. NIST reports that the NCCoE worked with 24 collaborators to build 19 example implementations. The guide offers practical ZTA examples; it does not establish a standardized code-execution control.
Recommended Free Tools
The distinction matters because Zero Trust Architecture generally addresses secure, authorized access to enterprise resources. Applying the same mindset to code means asking a different question: not only whether a user or workload may access a resource, but whether this code should be allowed to perform a particular action in the first place.
Rank #3
- POWERFUL SECURITY KEY: The YubiKey 5 NFC is the most versatile physical passkey, protecting your digital life from phishing attacks. It ensures only you can access your accounts
- WORKS WITH 1000+ ACCOUNTS: Compatible with popular accounts like Google, Microsoft, and Apple. A single YubiKey 5 NFC secures 100+ of your favorite accounts, including email, password managers, and more
- FAST & CONVENIENT LOGIN: Plug in your YubiKey 5 NFC via USB and tap it, or tap it against your phone (NFC), to authenticate. No batteries, no internet connection, and no extra fees required
- MOST SECURE PASSKEY: Supports FIDO2/WebAuthn, FIDO U2F, Yubico OTP, OATH-TOTP/HOTP, Smart card (PIV), and OpenPGP. That means it’s versatile, working almost anywhere you need it
- PRIMARY & SPARE KEYS: Just like having a spare house key, we recommend buying two YubiKeys - one for daily use and one as a spare. That way you’ll never get locked out of your accounts
Are SBOMs enough to keep software safe?
No. A software bill of materials (SBOM) is an inventory of software components that helps organizations understand supply chains and make risk-informed decisions. It can support vulnerability and dependency management, but listing ingredients does not show that a program’s runtime behavior is safe.
CISA and G7 partners’ May 12, 2026 guidance says AI-specific SBOM elements supplement general SBOM elements and describes them as non-exhaustive and non-mandatory. An SBOM is therefore one useful input to software risk decisions, not a runtime verdict. The same general caution applies to signatures and provenance: they can help establish integrity or origin, but do not by themselves guarantee safe behavior.
Rank #4
- POWERFUL SECURITY KEY: The Security Key NFC is the essential physical passkey for protecting your digital life from phishing attacks. It ensures only you can access your accounts.
- WORKS WITH 1000+ ACCOUNTS: Compatible with Google, Microsoft, and Apple. A single Security Key NFC secures 100 of your favorite accounts, including email, password managers, and more.
- FAST & CONVENIENT LOGIN: Plug in your Security Key NFC via USB-A and tap it, or tap it against your phone (NFC) to authenticate. No batteries, no internet connection, and no extra fees required.
- TRUSTED PASSKEY TECHNOLOGY: Uses the latest passkey standards (FIDO2/WebAuthn & FIDO U2F) but does not support One-Time Passwords. For complex needs, check out the YubiKey 5 Series.
- BUILT TO LAST: Made from tough, waterproof, and crush-resistant materials. Manufactured in Sweden and programmed in the USA with the highest security standards.
Why protect developer tools and CI/CD pipelines?
Developer environments and build pipelines can carry powerful access to source code, credentials, and release processes. CISA’s May 28, 2026 advisory describes two supply-chain compromises: a malicious Nx Console VS Code extension version distributed through automatic update, and malicious GitHub Actions workflows used to harvest secrets, credentials, and tokens. The advisory does not attribute either campaign to AI. They illustrate why the systems that create, update, and ship software need protection in their own right.
CISA recommends monitoring workflow files and contributor activity, pinning trusted versions, limiting package sources, and, after a compromise, investigating logs and affected machines and rotating credentials and tokens that were accessible. These measures reduce exposure in developer and CI/CD environments; they complement, rather than substitute for, checks on what code does when executed.
Best Value
- POWERFUL SECURITY KEY: The YubiKey 5 is a versatile physical passkey that protects your digital life from phishing attacks. It ensures only you can access your accounts.
- WORKS WITH 1000+ ACCOUNTS: Compatible with popular accounts like Google, Microsoft, and Apple. A single YubiKey 5 secures 100+ of your favorite accounts, including email, password managers, and more.
- FAST & CONVENIENT LOGIN: Plug in your YubiKey 5 via USB and tap it to authenticate. No batteries, no internet connection, and no extra fees required.
- MOST SECURE PASSKEY: Supports FIDO2/WebAuthn, FIDO U2F, Yubico OTP, OATH-TOTP/HOTP, Smart card (PIV), and OpenPGP. That means it’s versatile, working almost anywhere you need it.
- BUILT TO LAST: Made from tough, waterproof, and crush-resistant materials. Manufactured in Sweden and programmed in the USA with the highest security standards.
How can companies secure AI coding agents and code execution?
CISA and partners’ May 1, 2026 guidance, “Careful Adoption of Agentic Artificial Intelligence (AI) Services,” identifies risks including privilege escalation, emergent behaviors, and accountability gaps. It recommends limiting agent autonomy and access to sensitive data or critical systems, alongside layered defenses, strong identity management, oversight, threat modeling, continuous monitoring, and regular security assessments.
A practical program can turn those principles into a sequence of decisions. The behavior-policy step reflects Ammon’s proposal; the agent safeguards align with CISA’s guidance.
- Map where code enters and runs. Include repositories, dependencies, editor extensions, scripts, containers, CI/CD workflows, and code generated or invoked by agents. Record which identities and credentials each path can use.
- Identify inherited trust. Find places where origin, a signature, an approval, or a successful scan is treated as sufficient permission to execute. Keep those signals as evidence, but do not confuse them with proof of safe runtime behavior.
- Define allowed behavior and privilege. Specify what a program or agent needs to do, which resources it may access, and which actions should be denied. Apply the narrowest practical permissions to agents and their credentials.
- Make execution a policy decision. Where controls permit, evaluate code against expected behavior before it runs. Route exceptions for review or isolation rather than treating an unknown action as automatically trusted.
- Monitor execution and reassess. Log relevant activity, watch for changes to workflows and contributor behavior, and conduct regular security assessments. Use alerts and investigation to refine policies as systems and permissions change.
What should teams evaluate in a code-security control?
A useful evaluation focuses on what the control actually governs and when it acts—not just whether a product uses the words “Zero Trust.” The following questions help distinguish complementary safeguards:
Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Clear out junk files and repair common Windows errorsFree Scan →- What is controlled? Identity and resource access, software components and provenance, application behavior, or some combination?
- When does enforcement happen? At access time, before execution, or only after suspicious activity is detected?
- What is in scope? Does coverage include repositories, dependencies, extensions, scripts, containers, and CI/CD workflows, or only selected applications?
- How are agents constrained? Are autonomy, credentials, sensitive-data access, and access to critical systems limited and reviewable?
- Can decisions be audited? Can a team see what policy allowed or blocked an action and investigate the relevant execution?
These are evaluation criteria, not a vendor ranking or a claim that one control can replace the rest. The defensible goal is layered security: preserve useful supply-chain and development checks, constrain access, and add a deliberate policy check for runtime behavior where feasible.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
