Is Sift safe with your data?
Sift
Partial disclosure · grade C is 48–59 · high confidence
Scored 50 ± 5 (range 45–55)
That range crosses the C/D cut, so read this app as C or D rather than firmly either. How we grade
Sift earns a C (50/100) because it discloses its data practices only in part.
Dealbreaker flag
- D1.1: Sift may use the Customer Data for as long as reasonably necessary for the limited purpose of providing, maintaining, and improving the Sift Services, which includes improving its machine learning algorithms and data models
- D1.4: grants Sift an unlimited, irrevocable, perpetual, sublicensable, royalty-free license to use any such feedback or labels for any purpose
#91
of 216 apps ranked
50
score · Security & compliance avg 50
+0
vs category average
Sift tells business customers it will feed their data into its machine learning models and takes a perpetual, sublicensable licence over any labels they send back, while the website privacy notice covers individual rights, recipient categories and a ninety day deletion window but names no retention period, no security controls and no breach notification.
What Sift's privacy policy and terms of service say about your data
Customer data trains the global models
Sift states it may use customer data to improve its machine learning algorithms and data models, and that the data will be commingled with data from other customers. No opt-out is offered.
Perpetual licence over feedback labels
Customers who send fraud or abuse labels back to Sift grant an unlimited, irrevocable, perpetual, sublicensable, royalty-free licence to use them for any purpose.
Deletion has a timeline, retention has no number
After termination a customer can request deletion and Sift commits to ninety days, but the policy sets no retention period for ordinary use beyond an ongoing legitimate business need.
Security is described without naming controls
The notice cites technical and organizational security measures without naming encryption, access control or any certification, and it contains no breach notification commitment.
What the policy is silent or vague on
- Not stated: keeping user inputs out of model training
- Not stated: a way to opt out of training
- Not stated: whether training use differs by plan
- Not stated: your ownership of generated outputs
Sift privacy rating
Details
- Category
- Security & compliance
- Modalities
- text
- Processes biometrics
- No
- Policy last updated
- 2026-04-20
- Region scored
- Global / US-default
- Last assessed
- 2026-10-13
Documents examined
Each grade reflects our analysis of what an app states in its public privacy policy and terms as of the assessment date. It measures the transparency of those documents, not the company's actual data practices, security, or compliance. Grades are our opinion, offered for general information. Full disclaimer.