Worked Solutions · AP CSP-Feeder · US / ON / BC / AB Styles详细解答 · AP CSP 衔接 · 美 / 安 / 卑 / 阿省风格
Which type of malware disguises itself as legitimate software to trick a user into installing it, but then performs harmful actions once installed?哪种类型的恶意软件将自己伪装成合法软件，诱使用户安装，但安装后执行有害操作？
Correct: A Trojan disguises itself as legitimate software (e.g., a free game, a utility) to convince the user to install it voluntarily. Once installed, it performs malicious actions such as opening a backdoor, stealing data, or dropping other malware. It does not self-replicate (that is a virus) and does not spread across networks without a host file (that is a worm).正确：木马将自己伪装成合法软件（如免费游戏、实用工具），说服用户自愿安装。安装后，它执行恶意操作，如开后门、窃取数据或释放其他恶意软件。它不会自我复制（那是病毒），也不会在没有宿主文件的情况下在网络中传播（那是蠕虫）。
Distinguish between symmetric and asymmetric encryption; state which type is used to exchange a session key in HTTPS.区分对称加密和非对称加密；说明 HTTPS 使用哪种类型交换会话密钥。
Symmetric: 1 key (shared); Faster. | Asymmetric: 2 keys (public + private); Slower.对称：1 个密钥（共享）；更快。 | 非对称：2 个密钥（公钥 + 私钥）；更慢。
M1 Symmetric: 1 key / same key for encrypt and decrypt.对称：1 个密钥 / 加密解密使用相同密钥。
M1 Asymmetric: key pair (public + private); slower.非对称：密钥对（公钥 + 私钥）；更慢。
Asymmetric encryption is used to exchange the session key. M1 The reason is that symmetric encryption requires both parties to already share a key; if they have never communicated before, there is no safe way to share it over an untrusted channel. Asymmetric encryption solves this: the server publishes its public key; the browser uses it to send the session key securely; only the server's private key can decrypt it. A1非对称加密用于交换会话密钥。M1原因是对称加密要求双方已经共享密钥；如果他们从未通信过，就没有安全的方式在不受信任的信道上共享密钥。非对称加密解决了这个问题：服务器发布其公钥；浏览器用它安全地发送会话密钥；只有服务器的私钥才能解密。A1
Student uses "soccer2010" on 5 sites; one site is breached and hashes cracked.学生在 5 个网站上使用"soccer2010"；一个网站遭到入侵且哈希被破解。
Credential stuffing. A1 Attackers take a list of username/password pairs leaked from one site and automatically try them on many other sites (email, banking, social media). Because the student reused the same password, all five accounts are now at risk from one breach.凭据填充。A1攻击者获取从一个网站泄露的用户名/密码对列表，并在许多其他网站（电子邮件、银行、社交媒体）上自动尝试。因为学生重复使用了相同的密码，所有五个账户现在都因一次泄露而面临风险。
"soccer2010" is weak because: it is a common word (soccer) combined with a year (2010), making it a predictable pattern; it is only 10 characters; it uses no symbols or mixed case; and it is likely in common wordlists used by dictionary attacks. M1"soccer2010"很弱，因为：它是一个常用词（soccer）与年份（2010）的组合，形成可预测的模式；只有 10 个字符；没有使用符号或大小写混合；并且很可能在字典攻击使用的常用词汇表中。M1
Improvement 1: Increase length to at least 12-16 characters. A1改进 1：将长度增加到至少 12-16 个字符。A1
Improvement 2: Add symbols and mixed case, e.g., S0ccer!2O1O#x, or use a completely random passphrase. A1改进 2：添加符号和大小写混合，如 S0ccer!2O1O#x，或使用完全随机的密码短语。A1
Password manager (e.g., Bitwarden, 1Password, KeePass). A1 It generates a long, unique random password for every site and stores them securely; the user only memorises one master password.密码管理器（如 Bitwarden、1Password、KeePass）。A1它为每个网站生成一个长且唯一的随机密码并安全存储；用户只需记住一个主密码。
Navigation app requests location, microphone, contacts, and camera; stated purpose is turn-by-turn directions only.导航应用请求位置、麦克风、联系人和摄像头权限；声明目的仅为逐步导航。
Location (always on) is the only permission strictly necessary. A1 Turn-by-turn navigation requires knowing where you are at all times. Microphone, contacts, and camera have no functional relationship to providing directions. A1位置（始终开启）是唯一严格必要的权限。A1逐步导航需要随时知道你的位置。麦克风、联系人和摄像头与提供导航没有任何功能关系。A1
Data minimisation (also acceptable: principle of least privilege). A1 Definition: collect only the data strictly necessary for the stated purpose; do not gather data "just in case" or for unstated secondary uses. A1数据最小化（也可接受：最小权限原则）。A1定义：仅收集实现声明目的严格必要的数据；不要"以防万一"或为未声明的二次使用收集数据。A1
Tapping "Allow All" without reading does not constitute genuine informed consent because the user does not actually understand what data is being collected or why. M1 Informed consent requires that the user genuinely comprehends the tradeoff -- not merely clicks through a prompt to reach the app. A1未阅读就点击"全部允许"不构成真正的知情同意，因为用户实际上不了解正在收集什么数据或原因。M1知情同意要求用户真正理解权衡-而不仅仅是点击提示来进入应用。A1
Match scenarios to IP terms; explain one beneficial and one harmful effect of strict copyright on software innovation.将场景与知识产权术语匹配；解释严格版权对软件创新的一个有益影响和一个有害影响。
(a) Fair use. A1 Photocopying one chapter for classroom discussion without selling copies falls within educational fair use / fair dealing exceptions to copyright.(a) 合理使用。A1为课堂讨论复印一章内容且不出售副本，属于版权教育合理使用/公平交易例外。
(b) Software piracy. A1 Using proprietary software without a valid licence violates the software's copyright licence and constitutes piracy. Ontario ICS4U D2.1 lists "software piracy" as a key ethical issue.(b) 软件盗版。A1在没有有效许可的情况下使用专有软件违反了软件的版权许可，构成盗版。安大略 ICS4U D2.1 将"软件盗版"列为关键伦理问题。
(c) GPL (copyleft). A1 GPL's "viral" condition requires that any modified or derived work also be distributed under the GPL, ensuring the open-source ecosystem remains open. MIT licence is permissive -- it does not require derivatives to stay open.(c) GPL（著佐权）。A1GPL 的"传染性"条件要求任何修改或衍生作品也在 GPL 下分发，确保开源生态系统保持开放。MIT 许可是宽松的-它不要求衍生作品保持开放。
Beneficial (2 marks): Strict copyright gives creators and companies the legal right to profit from their software, which funds continued development, research, and maintenance -- without this incentive, much commercial software would not be created. M1 A1有益影响（2 分）：严格的版权给予创作者和公司从其软件中获利的法律权利，这为持续开发、研究和维护提供资金-没有这种激励，许多商业软件将不会被创建。M1 A1
Harmful (2 marks): Overly broad copyright protection can stifle innovation by preventing others from building on existing work (even to improve it), extending monopolies and slowing the collaborative community-driven progress that open-source enables. M1 A1有害影响（2 分）：过于宽泛的版权保护可能通过阻止他人基于现有作品进行开发（即使是为了改进）来抑制创新，延长垄断并减缓开源所促成的协作社区驱动进步。M1 A1
Identify threat category and state one defence for each scenario.识别每个场景的威胁类别并说明一种防御措施。
Threat: Phishing. A1 The attacker impersonates a trusted institution (the school) via email and uses a fake look-alike URL to steal credentials. The slight URL difference is the telltale sign of phishing. Defence: hover over links before clicking to verify the real URL; go directly to the school portal by typing the address; enable email spam/phishing filters. A1威胁：钓鱼。A1攻击者通过电子邮件冒充可信机构（学校），并使用虚假相似 URL 窃取凭据。略有不同的 URL 是钓鱼的标志。防御：点击前悬停在链接上验证真实 URL；通过键入地址直接访问学校门户；启用电子邮件垃圾/钓鱼过滤器。A1
Threat: Social engineering (pretexting / tech support scam). A1 No malware is required -- the attacker exploits fear and urgency to manipulate the user into granting access and paying money. Legitimate support services never cold-call users about infections. Defence: never call numbers from pop-ups; verify through the official vendor website; hang up on unsolicited "support" calls. A1威胁：社会工程学（借口攻击 / 技术支持诈骗）。A1不需要恶意软件-攻击者利用恐惧和紧迫感操纵用户授予访问权限并付款。合法的支持服务从不主动打电话给用户告知感染。防御：不要拨打弹窗中的电话号码；通过官方供应商网站验证；挂断主动打来的"支持"电话。A1
Threat: Malware (worm). A1 A worm spreads automatically across networks by exploiting software vulnerabilities -- no user interaction needed distinguishes it from a Trojan or phishing-delivered payload. Defence: apply OS patches and security updates promptly; use network segmentation and firewalls to limit lateral movement. A1威胁：恶意软件（蠕虫）。A1蠕虫通过利用软件漏洞在网络中自动传播-不需要用户交互，这将其与木马或钓鱼传递的有效载荷区分开来。防御：及时应用操作系统补丁和安全更新；使用网络分段和防火墙限制横向移动。A1
Caesar cipher shift 3; weakness; key exchange; three authentication factors.凯撒密码移位 3；弱点；密钥交换；三种身份验证因素。
H (8th letter) + 3 = K (11th). A (1st) + 3 = D (4th). C (3rd) + 3 = F (6th). K (11th) + 3 = N (14th).H（第 8 个字母）+ 3 = K（第 11 个）。A（第 1 个）+ 3 = D（第 4 个）。C（第 3 个）+ 3 = F（第 6 个）。K（第 11 个）+ 3 = N（第 14 个）。
Ciphertext: KDFN M1 (1 mark for correct method, 1 mark for correct answer) A1密文：KDFN M1（正确方法 1 分，正确答案 1 分）A1
The Caesar cipher has only 25 possible shift values, so an attacker can try all of them in seconds (brute force). M1 Additionally, it is vulnerable to frequency analysis: in English text, letters like E and T appear most often; matching the most common letters in the ciphertext to common English letters reveals the shift without even trying all keys. A1凯撒密码只有 25 个可能的移位值，因此攻击者可以在几秒钟内尝试所有值（暴力破解）。M1此外，它容易受到频率分析的攻击：在英文文本中，E 和 T 等字母出现最频繁；将密文中最常见的字母与常见英文字母匹配，即可在不尝试所有密钥的情况下揭示移位量。A1
Asymmetric encryption should be used to share the decryption key. A1 The reason: the sender can encrypt the symmetric decryption key using the recipient's public key; only the recipient's private key can decrypt it. This solves the key distribution problem -- there is no way to securely share a symmetric key over email without first having a shared secret. A1非对称加密应用于共享解密密钥。A1原因：发送者可以用接收者的公钥加密对称解密密钥；只有接收者的私钥才能解密。这解决了密钥分发问题-在没有预先共享秘密的情况下，无法通过电子邮件安全共享对称密钥。A1
Something you know (e.g., password, PIN); Something you have (e.g., phone, hardware key); Something you are (e.g., fingerprint, face scan). A1 (Award 1 mark for all three correct with examples.)你知道的东西（如密码、PIN）；你拥有的东西（如手机、硬件密钥）；你本身的特征（如指纹、面部扫描）。A1（三项均正确并有例子得 1 分。）
Grocery loyalty app: name, email, postal code, purchase history sold to data broker.杂货店会员应用：姓名、电子邮件、邮政编码、购买历史出售给数据经纪人。
Name is PII: it directly identifies an individual. A1姓名是 PII：它直接识别个人。A1
Email address is PII: it is a unique identifier linked to a specific person. A1电子邮件地址是 PII：它是与特定人员关联的唯一标识符。A1
Home postal code is PII: combined with name, it allows pinpointing where a person lives. A1 (Purchase history: accept as PII with justification -- detailed purchases can identify individuals via re-identification attacks; however, this is less directly PII than the first three.)家庭邮政编码是 PII：结合姓名，它可以精确定位一个人的居住地。A1（购买历史：有理由接受为 PII-详细购买记录可以通过重新识别攻击来识别个人；但是，这不如前三项直接属于 PII。）
The claim may be false because: detailed purchase histories can be re-identified even without names -- researchers have shown that as few as 3-4 purchases (item type, store, time) can uniquely identify a person within a large dataset. M1 A1 (Also acceptable: combining postal code with purchase patterns narrows down to very few individuals, enabling re-identification.)该声明可能是错误的，因为：即使没有姓名，详细的购买历史也可以被重新识别-研究人员已证明，仅 3-4 次购买（商品类型、商店、时间）就足以在大型数据集中唯一识别一个人。M1 A1（也可接受：将邮政编码与购买模式结合可以缩小到极少数个人，从而实现重新识别。）
Informed consent (or prohibition on secondary use without consent). A1 The data was collected for the purpose of processing loyalty rewards; selling it to third-party marketers is a secondary use that customers were not told about and did not agree to. This violates the principle that users must know and consent to how their data is used. A1知情同意（或禁止未经同意的二次使用）。A1数据是为处理会员奖励而收集的；将其出售给第三方营销商是客户未被告知且未同意的二次使用。这违反了用户必须知道并同意其数据使用方式的原则。A1
The GPL requires that the modified version of the library also be released under the GPL (distributed as open-source with the source code available). A1GPL 要求修改后的库版本也在 GPL 下发布（作为开源发布，源代码可用）。A1
AI hiring tool trained on 20 years of biased historical data; no screen-reader support.基于 20 年有偏见历史数据训练的 AI 招聘工具；无屏幕阅读器支持。
The problem is algorithmic bias (also: training data bias, or discriminatory AI). A1 If historical hiring data reflects patterns in which certain groups were systematically underrepresented or excluded (e.g., due to gender, race, or age bias), the AI will learn those patterns and reproduce them -- effectively automating discrimination at scale. A1问题是算法偏见（也称为：训练数据偏见，或歧视性 AI）。A1如果历史招聘数据反映了某些群体被系统性地低估或排除的模式（如由于性别、种族或年龄偏见），AI 将学习这些模式并复现它们-实际上是在大规模自动化歧视。A1
The argument is flawed. Identical treatment is not the same as equitable treatment: if the algorithm was trained on biased data, then treating all candidates "identically" by the same algorithm will reproduce the historical bias against the same groups. M1 A1 Equity requires that outcomes be fair across groups, not merely that the same process is applied; a process that systematically disadvantages a protected group is discriminatory regardless of whether it is "consistent." A1这个论点是有缺陷的。同样对待不等于公平对待：如果算法是在有偏见的数据上训练的，那么用同样的算法"同样地"对待所有候选人将会对相同群体复现历史偏见。M1 A1公平要求各群体的结果是公正的，而不仅仅是应用相同的过程；一个系统性地使受保护群体处于不利地位的过程是歧视性的，无论它是否"一致"。A1
Step 1: Conduct a bias audit before deployment -- statistically test the tool's outputs across demographic groups and compare acceptance rates; if disparate impact is found, retrain or adjust the model. A1措施 1：在部署前进行偏见审计-在不同人口群体中统计测试工具的输出并比较录取率；如果发现不公平影响，重新训练或调整模型。A1
Step 2: Curate and balance the training data -- actively include historical examples of successful employees from underrepresented groups (or use a dataset that does not reflect past discriminatory patterns), so the model learns from a more equitable baseline. A1措施 2：整理和平衡训练数据-主动纳入来自低估群体的成功员工的历史例子（或使用不反映过去歧视性模式的数据集），使模型从更公平的基线学习。A1
The portal violates digital accessibility (also: universal design, or inclusion in computing). A1 Without screen-reader support, visually impaired applicants cannot use the portal, meaning they are excluded from the application process entirely -- regardless of their qualifications. Since employment is a social and economic necessity, inaccessible digital systems compound existing inequities: they concentrate access among those who do not need assistive technologies, violating CSTA 3A-IC-30's requirement to consider the social and economic implications of computing on equity. A1该门户违反了数字可访问性（也称为：通用设计，或计算中的包容性）。A1没有屏幕阅读器支持，视觉障碍申请者无法使用该门户，这意味着他们完全被排除在申请过程之外-无论其资质如何。由于就业是一种社会和经济必需品，不可访问的数字系统加剧了现有的不平等：它们将访问集中于不需要辅助技术的人，违反了 CSTA 3A-IC-30 要求考虑计算对公平的社会和经济影响的规定。A1
Hospital ransomware attack via phishing email; no backups; 3-day disruption.医院通过钓鱼电子邮件遭受勒索软件攻击；无备份；中断 3 天。
Initial attack vector: phishing -- the attacker sent a deceptive email that the staff member opened, which either contained a malicious attachment or link. A1初始攻击向量：钓鱼-攻击者发送了工作人员打开的欺骗性电子邮件，其中包含恶意附件或链接。A1
Payload type: ransomware (a subtype of malware) -- the ransomware executed and encrypted all patient files, rendering them inaccessible. A1有效载荷类型：勒索软件（恶意软件的子类型）-勒索软件执行并加密了所有患者文件，使其无法访问。A1
Impact category: data unavailability / service disruption -- because patient files could not be accessed, patient care was disrupted for 3 days (also: potential data loss, reputational harm, financial cost of paying ransom or recovery). A1影响类别：数据不可用 / 服务中断-因为患者文件无法访问，患者护理中断了 3 天（也包括：潜在的数据丢失、声誉损害、支付赎金或恢复的财务成本）。A1
The claim that the staff member is "solely to blame" is incorrect -- the hospital's systemic failures amplified the impact significantly. M1关于工作人员"负有全部责任"的说法是不正确的-医院的系统性失误显著放大了影响。M1
Systemic failure 1: No recent backups. The 3-2-1 backup rule should have ensured offline copies of patient data; with backups, recovery is possible without paying the ransom. This is a hospital-level policy failure. A1系统性失误 1：没有近期备份。3-2-1 备份规则本应确保患者数据的离线副本；有了备份，无需支付赎金即可恢复。这是医院级别的政策失误。A1
Systemic failure 2: Insufficient staff security training. The hospital had not trained staff to identify phishing emails -- this is an organisational failure, not an individual one. Email filtering and phishing simulation training are standard healthcare security practices. A1系统性失误 2：员工安全培训不足。医院没有培训员工识别钓鱼电子邮件-这是组织失误，而非个人失误。电子邮件过滤和钓鱼模拟培训是标准的医疗安全实践。A1
Technical: Deploy an advanced email filtering system that detects and quarantines phishing emails before staff see them, reducing the risk of a single click becoming a breach. A1技术性：部署高级电子邮件过滤系统，在员工看到钓鱼电子邮件之前检测并隔离它们，降低一次点击成为泄露的风险。A1
Procedural: Implement a mandatory 3-2-1 backup policy with automated daily encrypted backups, including one offline or air-gapped copy, so that data can be restored without paying a ransom.程序性：实施强制性 3-2-1 备份政策，包括自动每日加密备份，其中包括一个离线或气隙副本，使数据可以在不支付赎金的情况下恢复。
Cultural: Conduct regular phishing simulation exercises (send fake phishing emails and track who clicks), followed by immediate targeted education for those who click, building a culture of scepticism about unexpected emails. A1 (Award A1 for any two well-justified measures; deduct if the same category is listed twice.)文化性：定期进行钓鱼模拟演练（发送虚假钓鱼电子邮件并追踪点击者），然后对点击者进行即时针对性教育，建立对意外电子邮件持怀疑态度的文化。A1（任何两项有充分理由的措施得 A1；如果同一类别列出两次则扣分。）
City proposes facial recognition cameras on public transit; privacy and bias concerns raised.城市提议在公共交通上安装人脸识别摄像头；提出隐私和偏见问题。
Category 1: Biometric data (facial geometry / facial recognition templates). A1 Facial geometry is PII because it uniquely identifies a person across contexts. Harm if breached: unlike a password, biometrics cannot be changed -- if the facial template database is stolen, affected individuals face permanent identification risk, enabling stalking, identity fraud, or targeting by malicious actors. A1类别 1：生物特征数据（面部几何 / 人脸识别模板）。A1面部几何是 PII，因为它在各种场景中唯一识别一个人。泄露后的危害：与密码不同，生物特征无法更改-如果人脸模板数据库被盗，受影响的个人面临永久识别风险，可能导致跟踪、身份欺诈或被恶意行为者针对。A1
Category 2: Location and movement history. A1 The system can reconstruct where each person traveled, when, and with whom. Harm if breached: location patterns reveal sensitive information -- visits to hospitals, religious sites, political gatherings, or homes of associates -- enabling surveillance, blackmail, or persecution of individuals based on where they go. A1类别 2：位置和移动历史。A1该系统可以重建每个人的出行地点、时间和同伴。泄露后的危害：位置模式揭示敏感信息-前往医院、宗教场所、政治集会或同伴住所-可能导致监视、勒索或基于行踪对个人的迫害。A1
The city should be required to obtain explicit public consent before deployment. The principle of informed consent holds that individuals must know what data is collected about them and agree to it before that collection occurs. M1 Deploying facial recognition in a public transit system -- which many residents have no practical alternative to -- means they are effectively compelled to submit to biometric surveillance without a genuine choice; this is not meaningful consent. A1 Without consent mechanisms (such as a public vote, legislative approval, or an opt-out for vulnerable groups), the city would be imposing surveillance on the public rather than serving them, undermining the trust that democratic governance requires. A1该市在部署前应被要求获得明确的公众同意。知情同意原则要求个人必须了解正在收集的数据并在收集发生前同意。M1在公共交通系统中部署人脸识别-许多居民没有实际替代方案-意味着他们实际上被迫在没有真正选择的情况下接受生物特征监控；这不是有意义的同意。A1没有同意机制（如公众投票、立法批准或弱势群体的退出选项），该市将对公众施加监控而非为其服务，破坏民主治理所需的信任。A1
ACM Code of Ethics (2018), Principle 1.6: "Respect privacy." A1 This principle directly applies because computing professionals who design or implement the facial recognition system have an obligation to consider and protect the privacy of the people the system will monitor; deploying a system with known privacy risks without adequate consent mechanisms would violate this principle. A1 (Also acceptable: IEEE Code of Ethics -- "to protect the privacy of others"; or ACM 3.6 on computing systems that serve the public interest.)ACM 伦理准则（2018），原则 1.6："尊重隐私。"A1这一原则直接适用，因为设计或实施人脸识别系统的计算专业人员有义务考虑并保护系统将监控的人员的隐私；在没有适当同意机制的情况下部署已知存在隐私风险的系统将违反这一原则。A1（也可接受：IEEE 伦理准则-"保护他人隐私"；或 ACM 3.6 关于服务公共利益的计算系统。）
Health records web app: plaintext passwords; selling anonymised data to insurers; MIT-licenced framework.健康记录网络应用：明文密码；将匿名化数据出售给保险公司；MIT 许可框架。
Security flaw: Storing passwords in plaintext means that anyone who gains access to the database (via SQL injection, insider threat, or a breach) can immediately read all passwords and use them on other sites (credential stuffing). M1安全缺陷：以明文存储密码意味着任何获得数据库访问权限的人（通过 SQL 注入、内部威胁或泄露）都可以立即读取所有密码并在其他网站上使用它们（凭据填充）。M1
Correct approach: Hashing with salting. A1 Instead of storing the password, store the output of a one-way cryptographic hash function (e.g., bcrypt, Argon2) applied to the password. A unique random string (the "salt") is appended to each password before hashing, ensuring that two users with the same password produce different hashes; this defeats precomputed rainbow table attacks. When the user logs in, the server hashes the entered password with the same salt and compares the result to the stored hash. A1正确做法：带加盐的哈希。A1不存储密码，而是存储应用于密码的单向加密哈希函数（如 bcrypt、Argon2）的输出。在哈希之前向每个密码附加一个唯一的随机字符串（"盐"），确保具有相同密码的两个用户产生不同的哈希；这可以抵御预先计算的彩虹表攻击。用户登录时，服务器使用相同的盐对输入的密码进行哈希，并将结果与存储的哈希进行比较。A1
PII 1: Medical history. Security measure: encrypt this data at rest (using AES) so that even a database breach reveals only ciphertext, not readable health records. A1PII 1：病历。安全措施：对此数据进行静态加密（使用 AES），使即使发生数据库泄露也只暴露密文，而非可读的健康记录。A1
PII 2: Date of birth. Security measure: implement strict access controls (role-based access) so that only authorised users (e.g., the patient's own doctor) can query date of birth; limit exposure by displaying only the minimum required fields in interfaces. A1PII 2：出生日期。安全措施：实施严格的访问控制（基于角色的访问），使只有授权用户（如患者自己的医生）才能查询出生日期；通过在界面中仅显示所需的最少字段来限制暴露。A1
Ethical violation: Selling user data to health insurers without informed consent (secondary use without consent; also: breach of data minimisation). A1伦理违规：在未获得知情同意的情况下将用户数据出售给健康保险公司（未经同意的二次使用；也：违反数据最小化）。A1
Anonymisation alone does not protect users in this context for two reasons: M1 First, health data is highly specific -- rare conditions, treatments, and prescription patterns can re-identify individuals even without names, particularly in the context of a small geographic area. Second, and more critically for this specific scenario: health insurers are highly motivated to re-identify individuals because doing so allows them to deny coverage, raise premiums, or discriminate based on pre-existing conditions. The incentive structure for the buyer actively works against anonymisation holding. A1在这种情况下，单纯匿名化无法保护用户有两个原因：M1首先，健康数据高度具体-即使没有姓名，罕见病症、治疗和处方模式也可以重新识别个人，特别是在小地理区域的情况下。其次，对于这个具体场景更为关键：健康保险公司高度有动力重新识别个人，因为这样做使他们能够拒绝保险、提高保费或基于既往病症进行歧视。买方的激励结构积极地反对匿名化的维持。A1