# Attribution Reality Check Kit Purpose: a printable set of worksheets for reading breach news without overstating what is known. Use it when a vendor-script compromise (or any incident) hits the headlines and you need to brief staff, answer customers or review your own website. It separates confirmed fact from reported claim from attribution, and points you back at the defensive steps that matter whoever is blamed. ## Checklist - [ ] Write down the incident in one sentence using only what the vendor or victim has confirmed. - [ ] Open at least two original sources (not just summaries) and note the outlet and date of each. - [ ] Copy every factual sentence into one of three columns: confirmed, reported, attributed. - [ ] For each reported claim, note whether the outlet has its own reporting or is repeating another. - [ ] Count independent sources (vendor statement, researcher write-up and news report count once each; repeats count as one). - [ ] Find every sentence naming a country, group or person-type actor and mark it "attribution". - [ ] Ask: what published evidence (indicators, root cause, named authority) would support that attribution? - [ ] Ask: would my actions change depending on the answer? If no, proceed with the defensive steps anyway. - [ ] List every third-party script on your site and which pages it loads on. - [ ] Remove or pause any script you don't actively use, starting with checkout, login and form pages. - [ ] Ask your developer or host about a Content-Security-Policy in report-only mode. - [ ] Email each relevant vendor asking whether your account or pages were affected. - [ ] Write a staff or customer note using only calibrated wording (confirmed, reported, we believe, we have no evidence of). - [ ] Set a re-check date and fill in the "what would change my mind" log. ## Worksheet 1: three-tier sort Filled example (Optimizely story, from public reporting as summarised by Apprised.news's tech and cyber desk, 2026-11-18): | Tier | Statement | Source | Independent? | |---|---|---|---| | Confirmed | Optimizely confirmed the breach | Vendor, via CSO Online | Yes | | Reported | At least 178 top sites served malware exploits in about 24 hours | Krebs on Security | Yes | | Attributed | "Apparently from China" | Stated plainly by one outlet; treated by the desk as an unsupported lead | Unresolved | Your blank copy: | Tier | Statement | Source | Independent? (yes / no / unsure) | |---|---|---|---| | Confirmed | | | | | Confirmed | | | | | Reported | | | | | Reported | | | | | Attributed | | | | | Attributed | | | | ## Worksheet 2: source-independence checklist For each article or post, answer: - [ ] Does it name its own sources or evidence? - [ ] Does it link or credit another outlet for the key claim? - [ ] Is it a vendor or victim statement, a researcher write-up or a news report? (Each counts once.) - [ ] Does it use hedges ("apparently", "reportedly") or state the claim flatly? - [ ] Does any named authority put its name to the attribution? - [ ] Are indicators or a root cause published anywhere? Total independent sources: ______ ## Worksheet 3: confidence-language cheat sheet | Situation | Say this | Not this | |---|---|---| | Vendor admitted it | "The vendor has confirmed a breach." | "We were hacked." (if you don't know yet) | | One outlet reports it | "It has been reported that..." | "It is known that..." | | Actor named, no evidence | "One outlet attributes this to X; that is not independently confirmed." | "X did this." | | No evidence either way | "We have no evidence either way about who is responsible." | "It was probably X." | | Waiting on disclosure | "We will update when the vendor publishes its root cause." | "The investigation is complete." | ## Worksheet 4: third-party script inventory | Script / vendor | Pages it loads on | Runs where customers type data? (Y/N) | Who can change it without our approval? | Would we know within 24 hours if it changed? (Y/N) | Risk (high/med/low) | Action | |---|---|---|---|---|---|---| | | | | | | | | | | | | | | | | | | | | | | | | Rule of thumb: "yes" to data pages, "anyone at the vendor" for changes and "no" to noticing changes goes to the top of the review list. ## Worksheet 5: "what would change my mind" log | Date | Claim I am holding | Evidence that would change my view | New evidence seen | Updated view | |---|---|---|---|---| | | | | | | | | | | | | ## Template: vendor question email ```text Subject: Request for incident information: [vendor name] Hello, We use [product/script] on [pages]. Following recent public reports of a compromise involving your service, please tell us in writing: 1. Whether our account, site or pages were affected, and during what dates and times. 2. Whether any script served from your systems to our visitors was altered, and how you know. 3. What data, if any, could have been exposed to our visitors. 4. Whether you have published or will publish a root cause, and where. 5. What we should change on our side (keys, passwords, script settings). Please send your formal incident notice as soon as it is available. Thank you, [name, role, contact] ``` ## Template: customer reply ```text Thanks for asking. A third-party script vendor was compromised for about 24 hours, and some websites were affected. We are checking whether our site used that vendor and what it loaded, and we will update you by [date]. We have no evidence either way about who is responsible, and we are not going to guess. If you see an unfamiliar charge, please contact your card issuer. ``` ## Prompts (ready to paste; add your own documents) Sort claims: ```text Here are news snippets and statements about a vendor-script breach: [paste]. Sort every claim into three buckets: confirmed by the vendor or victim, reported by outlets, attributed to a named actor. Flag claims that share a single original source. If you cannot tell which bucket a claim belongs in, say so instead of guessing. ``` Skeptical analyst: ```text Act as a skeptical analyst. Given this attribution claim and context: [paste], list what evidence would support it, what would weaken it, and suggest careful confidence wording for a customer notice. Tell me when the text is too thin to judge. ``` Script risk plan: ```text Here is my list of third-party scripts and the pages they load on: [paste]. Rank by risk if the vendor were compromised and draft a plain-English Content-Security-Policy plan for my web developer, starting in report-only mode. Mark anything you cannot identify from my list. ``` Customer statement: ```text Draft a customer-facing statement of about 120 words about a vendor-script incident using only these facts: [paste]. Say what is confirmed, what we are doing and when we will update. Do not speculate about who is responsible. Flag anything in my facts that sounds like a guess. ``` Competing hypotheses: ```text Walk me through competing hypotheses for who or what caused the incident in my notes: [paste]. List at least four, the evidence each would predict, and the single disclosure that would best separate them. Say where you are unsure. ``` ## Sources - https://apprised.news/desk/tech/2026-11-18 - https://krebsonsecurity.com/2026/11/attackers-used-optimizely-to-a-b-test-malware-on-top-websites/ - https://www.csoonline.com/article/4249925/optimizely-confirms-massive-cyberattack-on-customer-websites-injected-stealer-malware.html - https://unit42.paloaltonetworks.com/optimjack-supply-chain-attack/