Lori Schafer in Forbes Tech Council: Strengthen Data Partnerships With Proven Privacy Safeguards

Read the original article in Forbes Technology Council here.

More B2B companies are leaning into data-sharing partnerships to sharpen decision-making and streamline operations. While the approach can open up valuable opportunities, it’s essential to remember that privacy and trust matter as much as performance.

Clear guardrails, aligned expectations and well-chosen tools can turn these collaborations into reliable and secure engines of value. Below, members of Forbes Technology Council share practical tips for structuring partnerships that unlock meaningful insights without putting critical information at risk.

Redact Or Mask The Data

Wherever possible, don’t share the private data. Redact it or mask it—either in the local systems or on the fly during transit to the collaboration system. This is the strongest way of protecting sensitive data, but it does require an investment in building the redaction capability. - Jeff Torello, Sinjun AI

Keep Data In Controlled Systems

It is imperative to not allow data of this nature to hit personal devices or email accounts. Keep the data in controlled systems (even generic ones like SharePoint) and prohibit downloads and exfiltration. Combine that with audit logging to know what has been accessed and when, as well as who changed what, and you’re all set to collaborate safely. - Martin Snyder, Certara

Clarify Which Applications And Services Will Manage The Data

In addition to regular contractual clauses covering “what” the data will be used for, insist on explicit clauses that dictate the “applications and services” in which the data will be stored and processed. Not all applications are equal; our recent research shows 84% of AI apps don’t encrypt data at rest, for example. Do your own research on the proposed systems before agreeing to share any data. - Steve Tait, Skyhigh Security

Prioritize Real-Time Workload Visibility

The key is to focus on comprehensive, real-time workload visibility. Think of it like this: You need a live dashboard that shows exactly what data is being used, where it is and what AI is touching it. This visibility helps your teams quickly see and fix potential privacy issues, ensuring you can keep the data flowing for great business insights while securely following all the rules. - Guillaume Aymé, Lenses.io

Embed Strong Privacy And Security Standards At Every Stage

You need to establish clear governance from the start, especially when financial data is involved. Define what’s shared, how it is used and who owns the insights. Embedding rigorous privacy and security standards at every stage builds the trust needed to unlock value without compromising sensitive financial information. - Michael Praeger, AvidXchange

Adopt A Zero-Trust Model

One effective way to preserve privacy, security and trust is to adopt a zero-trust data exchange model. Instead of assuming partners or systems are inherently trustworthy, every access request, dataset and API interaction should be continuously verified based on identity, context and risk. - Rohit Shirwadkar, Equinix

Create A ‘Data Clean Room’ For Partnerships

Implement a “data clean room.” This neutral, secure environment allows partners to pool and analyze encrypted data for collective insights without ever exposing raw customer information to each other. It replaces trust with cryptographic proof, ensuring privacy compliance and enabling high-value collaboration. - Miguel Llorca, Axazure

Design Trust By Architecture

The key to effective B2B data partnerships is designing trust by architecture. Beyond contracts, partners should establish shared governance, encryption-in-use and clear data lineage. Data must flow through secure, auditable layers where access is transparent, yet privacy-preserving. The goal isn’t just compliance—it’s mutual accountability that turns data sharing into a real advantage. - Regan Peng, PINAI

Share Insights, Not Raw Data

Focus on sharing insights and conclusions, not data. Use techniques like differential privacy or secure enclaves where computation can happen on encrypted data in isolation. Partners get answers instead of access. If raw data has to leave your environment, you’re negotiating a hostage situation, not a partnership. - JC Grubbs, Tandem

Use Transparent Data Layers And Shared Lineage Tracking

Drawing from my experience designing secure multiteam data ecosystems, I’ve learned that data sharing should be like building a bridge, not a tunnel. Both sides should see the flow, not just the outcome. Transparent data layers and shared lineage tracking let partners collaborate confidently. Value moves freely, but control never leaves either side. - Vivek Venkatesan, The Vanguard Group

Grant Data Access On An As-Needed Basis

Clearly defining roles and granting sensitive data access only to those who truly need it is one excellent practice. Get partners’ honest consent, and use robust encryption. This fosters confidence, protects data and facilitates seamless collaboration while safeguarding critical information. - Jay Krishnan, NAIB IT Consultancy Solutions WLL

Develop Unified, Permissioned Data Models For Collaboration

Adopt a governed data-sharing architecture built on a single source of truth. When all partners collaborate through unified, permissioned data models, you enable transparency, accuracy and AI-ready insights without exposing sensitive information. True value comes from connecting intelligence, not just exchanging data. - Lori Schafer, Digital Wave Technology

Plan For The End With A ‘Data Pre-Sharing Prenup’

As with marriages, not all data-sharing collaborations last forever. As such, companies should always plan for the eventual end of the partnership. Create a “data pre-sharing prenup” agreement to determine long-term data ownership, usage rights, privacy and security. The data ownership fiasco from the 23andMe bankruptcy highlights the importance of proactively considering these issues. - Mark Francis, Electronic Caregiver

Adopt ‘Consensus Vaulting’ To Achieve Cryptographic Proof Of Trust

Adopt “consensus vaulting”: Partners encrypt data within a neutral enclave governed by smart contracts. Algorithms run only when all digital signatures validate intent, producing shared intelligence without exposing source data. In AI-driven ecosystems such as supply chains, this transforms compliance from paperwork into cryptographic proof of trust. - Dileep Rai, Hachette Book Group

Follow The ‘Trust But Verify’ Model

Our data partnerships follow a “trust but verify” model. Each is guided by a tailored risk assessment and a shared security responsibilities agreement that defines partner-specific controls, ensuring collaboration drives value without compromising data protection. - Jane Mason, Clarifire

Align On Security Frameworks

Start by aligning on established security frameworks like SOC 2 or ISO 27001. These standards define clear controls for protecting sensitive data and require your partners to uphold the same practices. This shared compliance foundation builds trust and ensures data collaborations remain secure and valuable. - Nikita Fedorov, Qase.io

Ensure Information Contributors Retain Ownership And Earn Value

The strongest data partnerships treat information as a shared, tokenized asset that generates measurable value for all participants. Structure agreements so each contributor retains traceable ownership and earns value as insights are produced. Use encrypted, permissioned systems that log every exchange transparently. When protection and performance align, collaboration becomes truly profitable. - Charles Morey, MobilEyes Inc.

Co-Design Data Sharing Protocols

From my experience working across data-heavy B2B environments, one best practice that consistently drives value is co-designing data-sharing protocols with partners. When both sides define sensitivity levels and access rules together, it not only protects privacy, but also builds trust and ensures the data is truly actionable. - Laxmi Vanam

Build ‘Value Firewalls’—Share Algorithms, Not Raw Data

Build “value firewalls”—define not only how data is protected, but also how value is shared. Structure partnerships where algorithms, not raw data, travel across boundaries. This lets each party contribute intelligence safely, creating a trust economy where insights flow freely while ownership and privacy stay intact. - Ajit Prakash, ZENSAR TECHNOLOGIES LTD

Treat Partnerships As ‘Use-Case Contracts’

Treat data partnerships as “use-case contracts.” Codify purpose, fields and retention requirements in a shared schema, then enforce those terms with policy-as-code, clean-room execution and signed audit trails. When possible, move compute to the data through approaches such as federated learning or trusted execution environments (TEE), and share scored outputs or synthetic cohorts—not rows. We have found that this pattern helps us unlock signal while keeping data exposure provably bounded. - Rohit Anabheri, LotusPetal AI