Share
Embed
Slide 1
Slide 1 text
Privacy on Rails
Pragmatically complying to data protection laws
Slide 2
Slide 2 text
Talysson Oliveira
Software architect
& Chief Learning Officer
at Codeminer42
@talyssonoc
beacons.ai/talyssonoc
Slide 3
Slide 3 text
codeminer42.com
ai.codeminer42.com
@codeminer42
Slide 4
Slide 4 text
⚠
I'm not a lawyer
This is an engineering perspective talk
Slide 5
Slide 5 text
On calm and peaceful day…
Slide 6
Slide 6 text
No content
Slide 7
Slide 7 text
Would you know what to do?
Slide 8
Slide 8 text
Or would you freak out thinking…
- What tables store user private information?
- Are we able to remove them from the logs too?
- Oh no, what about ActiveJob payloads?
- How much time do we have to respond to that email?!
- Are we breaking any law? Are we going to be fined?
Slide 9
Slide 9 text
What are data protection laws?
Slide 10
Slide 10 text
Data protection laws
- Exist to enforce data protection and user privacy
- Present in more than 140 countries
- Most don't apply only if you are based in the country, but also if you
have users there
- They are about consent, data minimization, access rights, breach
notification, and purpose limitation
Slide 11
Slide 11 text
What private data are we talking about?
Slide 12
Slide 12 text
Name
Email
Phone
Government ID
Date of
birth
Race
Religion
Prescriptions
Tax records
IP address
Browsing
history
Chat
messages
Salary
Behavior
profile
AI generated
predictions
Gender
identity
Location
history
Slide 13
Slide 13 text
Name
Email
Phone
Government ID
Date of
birth
Race
Religion
Prescriptions
Tax records
IP address
Browsing
history
Chat
messages
Salary
Behavior
profile
AI generated
predictions
Gender
identity
Location
history
AND MORE!
Slide 14
Slide 14 text
Thinking about it made you sweat?
Then this talk's for you
Slide 15
Slide 15 text
Privacy used to be a competitive advantage
Now, it's the baseline for a product
Slide 16
Slide 16 text
Why do companies fail to comply?
Slide 17
Slide 17 text
They treat privacy as just a legal concern
They put innovation above user privacy
Slide 18
Slide 18 text
Privacy by design
Privacy as an engineering quality attribute
Slide 19
Slide 19 text
Privacy by design
- Don't implement privacy as an afterthought
- Implement it as any other normal engineering constraint
- Like we do for security, for performance, for responsiveness
- Think about your user's privacy first
- Don't gamble with user data to pay for innovation
- Start with a set of principles
Slide 20
Slide 20 text
Privacy by design principles
1. Data minimization
- Don't collect what you can't protect (or don't need)
- Less data = less to concern about
2. Private by default
- Every data exposure requires an explicit reason
- Data must be protected with appropriate technical measures
3. Transparency
- Users should know and control what you store from them
Slide 21
Slide 21 text
Privacy on Rails
"I'm not a lawyer either, show me the code!"
Slide 22
Slide 22 text
Data minimization
Less data = less to concern about
Slide 23
Slide 23 text
Minimize data that enters your app
Slide 24
Slide 24 text
Use strong parameters
- Create an explicitly permitted list of parameters
- Be intentional about what you permit
- Filtering at the entry point covers the whole application
Slide 25
Slide 25 text
Anonymize user's IPs
- Doing it at the entry point covers the whole application
- If geocoding is important for your app, it'll still work
- Use the ip_anonymizer gem
Slide 26
Slide 26 text
Minimize data that leaves your app
Slide 27
Slide 27 text
Explicit serialization
- Create an explicitly permitted list of what data will be exposed
Slide 28
Slide 28 text
Safeguard error tracking strategies
- Error tracking services capture full request context (PII, params, …)
- Filter request params and use the logstop gem
Slide 29
Slide 29 text
Anonymize analytics data
- Innovation should not be at the expense of your user's privacy
- Use a privacy-first analytics service + consent-gated scripts
- Anonymize any personal information sent do analytics
- We'll see consent modeling and anonymization techniques soon
Slide 30
Slide 30 text
No personal data in emails
- Data can leak in email logs and bounced email handling
- Enforce minimal PII in emails and filter email logs
- Use signed URLs with auto-expiry and authenticated downloads
Slide 31
Slide 31 text
Data retention and TTL policies
Slide 32
Slide 32 text
Implement data retention and TTL policies
- Don't keep older unused data
- Delete (or anonymize) inactive users
- Delete inactive sessions
- Clear SolidQueue completed job records
- Make it automatic
Slide 33
Slide 33 text
Deleting/anonymizing inactive users
Slide 34
Slide 34 text
Private data by default
Protected data without extra effort
Slide 35
Slide 35 text
Logs
Slide 36
Slide 36 text
Log with care
- Logs are the #1 silent leak
- Always check everything that is explicitly logged
- Filter request params and use the logstop gem
Slide 37
Slide 37 text
Security
Slide 38
Slide 38 text
Encryption at rest for models
Decrypt Search Use
✅ ✅
✅ ❌
❌ ❌
(just verify)
- Rails supports three levels of data protection
- Choose based on the field's usage pattern
Slide 39
Slide 39 text
Protect cookies properly
- Be intentional about cookies usage
- Define proper same_site policy
Slide 40
Slide 40 text
Make your backups secure
- Always encrypt the database backups
- Adopt a retention policy for backups as well
Slide 41
Slide 41 text
Protect direct console access in production
- Track who, why and how console was accessed in production
- Use console1984 and audits1984 gems for that
Slide 42
Slide 42 text
Transparency and data rights
Users should know and control their data
Slide 43
Slide 43 text
Consent
Slide 44
Slide 44 text
Creating a consent model
- It's not just about using a cookie banner
- Marketing, analytics tracking, 3rd party sharing, features, …
- Each consent should be per-purpose, versioned and with proof
- Consents should be explicit
- No user consent = no action
Slide 45
Slide 45 text
Creating a consent model
Slide 46
Slide 46 text
No content
Slide 47
Slide 47 text
Data subject access rights
Slide 48
Slide 48 text
Data subject access rights
- Users have access, rectification and erasure rights over their data
- LGPD requires a response in 15 days to respond
- GDPR requires a response in 30 days (extendable until 90 days)
- Make it as automatic as possible
Slide 49
Slide 49 text
Creating a data subject request model
Slide 50
Slide 50 text
Implementing access request (aka DSAR) handling
Slide 51
Slide 51 text
Implementing access request (aka DSAR) handling
Slide 52
Slide 52 text
Implementing erasure request handling
Slide 53
Slide 53 text
Implementing anonymization
Slide 54
Slide 54 text
Implementing anonymization
Slide 55
Slide 55 text
Ok, this is a lot of things to remember
Slide 56
Slide 56 text
And we are in the age of AI…
Slide 57
Slide 57 text
Automate it with agent skills!
Slide 58
Slide 58 text
Privacy by design Rails agent skills, by Codeminer42
- Bring privacy-by-design into your Rails workflow
- Two modes:
- Complete codebase assessment
- Review your recent changes
- Generates a thorough report + offer fixes
- Fully open source and ship with Ruby scripts for easy auditing
Slide 59
Slide 59 text
Privacy by design Rails agent skills, by Codeminer42
or
https://github.com/codeminer42/skills
Slide 60
Slide 60 text
Privacy by design Rails agent skills, by Codeminer42
Slide 61
Slide 61 text
Privacy by design Rails agent skills, by Codeminer42
Slide 62
Slide 62 text
Privacy by design Rails agent skills, by Codeminer42
Slide 63
Slide 63 text
Privacy by design Rails agent skills, by Codeminer42
Slide 64
Slide 64 text
Data privacy laws don't kill innovation
Not complying is a skill issue
Respect your users privacy
Rails is privacy friendly
Slide 65
Slide 65 text
Thank you!
@talyssonoc
beacons.ai/talyssonoc
codeminer42.com
ai.codeminer42.com
@codeminer42