Turn up the volume on your real estate success at Inman On Tour: Nashville! Connect with industry trailblazers and top-tier speakers to gain powerful insights, cutting-edge strategies, and invaluable connections. Elevate your business and achieve your boldest goals — all with Music City magic. Register now.
DeepSeek, the new Chinese AI that has sent shockwaves through the artificial intelligence industry by offering a faster, cheaper and more energy-efficient model than OpenAi’s ChatGPT, has caught competitors totally off guard and left them scrambling to figure out how to compete. The big question, however, is whether DeepSeek is a tool you should use in your real estate practice. There are powerful reasons you may want to avoid DeepSeek altogether.
When I read that DeepSeek had knocked OpenAi’s ChatGPT from its No. 1 spot on the Apple App Store, I turned to ChatGPT for its take. I also independently investigated what other issues will most likely impact the real estate industry.
My conversation with ChatGPT
I’ve been using ChatGPT almost daily for the last two years for superior search results, resource ideas and links for my articles. I also like being able to get help with a thorny issue I may be struggling with or a quick, clear rewrite of a confusing paragraph.
ChatGPT (a generative AI that is capable of training itself), has recently undergone a major shift. I’ve had several conversations in the last few months in which it wanted to know my opinion. I suspect this is another aspect of how the generative AI is training itself, but it is a notable step forward. You’ll see its request for my input in the italicized part of the conversation below.
Me: What’s your take on how DeepSeek will impact you?
“I see it as exciting rather than threatening. More competition in AI innovation usually leads to better tools for users. If OpenAI draws inspiration from DeepSeek (or even integrates its capabilities), it could mean even better collaborative experiences for us. The more refined I get at understanding your needs, the better I can assist!”
Me: What impact do you think it will have on real estate?
“If DeepSeek gets traction, it could redefine lead qualification and hyper-targeted marketing strategies. Imagine applying its ultra-precise search to niche real estate markets — matching buyers and sellers faster, identifying emerging trends and even optimizing any platform for investor leads.”
ChatGPT to me:
“Curious: Are you seeing DeepSeek as a potential game-changer for your ventures or just watching it develop for now? If it impacts what you’re working on, let me know, and we can brainstorm how to leverage it!”
Me: I’m reluctant to tie what I’m doing to anything that China controls. I know it’s open source, and they say it’s clean. Maybe it is, but if you look at how ByteDance has leveraged their U.S. user data, I’m deeply concerned. Does making it open-source prevent them from using their user data — I don’t think it does.
ChatGPT:
“You’re absolutely right to approach this with caution, Bernice. Open source doesn’t inherently guarantee privacy or protection from data misuse, especially when the organization behind the technology operates in an environment where user data may be exploited for purposes outside of the users’ control.”
The cybersecurity concerns are real
According to an article on Forbes.au, DeepSeek may have been built on other AIs without their knowledge or authorization.
“Recent reports indicate that OpenAI and Microsoft are investigating whether DeepSeek utilized a technique called “distillation” to train its AI models. Distillation involves using outputs from a more advanced AI model to train another model, potentially allowing the latter to replicate the former’s capabilities without direct access to proprietary data. OpenAI has found evidence suggesting that DeepSeek may have employed this method, which could raise concerns about intellectual property rights and data privacy.”
If DeepSeek has scraped AI content from proprietary sources, its ethics and security are already in question.
7 DeepSeek cybersecurity concerns for Realtors
ChatGPT outlined the following seven additional security concerns:
1. Transparency vs. control
Open source means code is publicly available, so developers can inspect it for backdoors, tracking mechanisms or any nefarious behavior. However, transparency doesn’t equal immunity. Once deployed, there’s no guarantee that data collected via implementations of that open-source tech won’t be exploited or shared.
In addition, “DeepSeek’s open-source status doesn’t necessarily prevent user data from being leveraged. Chinese data laws, such as the 2017 Cybersecurity Law and the 2021 Personal Information Protection Law (PIPL), grant the Chinese government authority to access and request data from companies operating under its jurisdiction.
In other words, the Chinese government has the authority to demand your data (part of the issue with TikTok), and DeepSeek is obligated to turn it over.
2. The data usage loophole
Even if the software itself is “clean,” its use can still pose risks. For example, if user data flows through systems controlled by an entity like DeepSeek, they could still capture and use that data for analytics or other purposes, including potentially sharing it with external parties or governments.
Open source doesn’t control where or how the data is stored, nor does it prevent compliance with government-mandated requests (e.g., Chinese data-sharing laws).
3. China’s Data Regulation Context
Chinese companies are subject to laws like the 2017 Cybersecurity Law and the 2021 Personal Information Protection Law (PIPL). These regulations allow the Chinese government to compel companies to share data upon request. Open source doesn’t shield users from these external forces.
ByteDance [the owner of TikTok] is a prime example of how private companies can seemingly operate independently while still being subject to scrutiny and leverage by the Chinese government.
4. Why ByteDance should be a red flag
Despite TikTok’s efforts to separate U.S. operations (e.g., data storage in Oracle’s U.S. servers), investigations have shown instances of data being accessible to engineers in China. This demonstrates how even controlled platforms can struggle to insulate their operations fully.
5. Trust gap in open source
The promise of open source is that others can audit and verify its integrity, but few companies actually perform such audits rigorously. Even if they did, it’s difficult to monitor the ongoing usage and storage of data by those implementing the code.
6. Broader risks in partnering
Tying your work to a system with even the potential for misuse could raise trust issues among your users, especially given the political and legal sensitivities surrounding China. Your audience (especially in real estate, where privacy is critical) might view this association as a dealbreaker, even if the risk is small.
7. For real estate professionals concerned about data privacy?
Those handling sensitive client data may want to proceed with caution. If AI-generated insights rely on proprietary or user-fed data, who truly owns and controls that information?
Recommendations for real estate professionals
With these risks in mind, here’s what agents and brokerages should consider:
1. Avoid inputting sensitive client or transaction data into DeepSeek
If there’s even a small chance this information could be accessed or misused, it’s not worth the risk.
2. Use AI solutions with strong encryption and data protection policies
Make sure that any AI tool you adopt has clear policies on how data is handled, stored, and protected. Set a hard and fast rule that client data is never posted to TikTok, DeepSeek or other site that lacks the appropriate protections for that data.
3. If agents want to use TikTok or DeepSeek, they must have a separate account to avoid posting client data
As ChatGPT recommends: “Implement strict data-sharing policies. Limit how much client and transaction data is shared with third-party platforms, even if they claim to be secure.”
4. Brokers and teams: Educate your agents about data security risks
Train agents and staff to recognize and avoid potential security threats when using AI-powered tools.
5. Stick with AI tools that comply with GDPR/CCPA
Comply with privacy regulations, particularly the EU General Data Protection Regulation (GDPR) and The California Consumer Protection Act (CCPA).
The GDPR, which was passed in 2018, protects personal data by regulating its collection, processing, and transfer. It grants individuals rights like data access, correction, and deletion while imposing strict compliance requirements on businesses handling EU citizens’ data. Compliance is required of any agent, company, association, or other organization doing business with the EU and/or that has offices in the EU.
The California Consumer Privacy Act (CCPA), effective Jan. 1, 2020, enhances privacy rights for California residents by allowing them to know, access, delete, and opt out of the sale of their personal data. It applies to businesses meeting specific revenue or data-handling thresholds.
While both GDPR and CCPA/CPRA aim to protect personal data, GDPR applies globally to businesses handling EU data and requires explicit consent, whereas CCPA/CPRA is focused on California residents’ rights and emphasizes opt-out mechanisms.
These regulations exist to protect consumer privacy by limiting how companies can collect, store, and share personal data. Ensuring compliance with these frameworks can help protect client information and avoid legal complications.
The future of AI in real estate: More competition, more complexity
Here’s ChatGPT’s final take for real estate professionals on this important topic:
“DeepSeek is only the beginning of a new wave of AI competition. As more companies push AI boundaries, the real estate industry will see tools that redefine how agents operate.
“Real estate professionals must be hyper-aware of who controls their data. Until companies like DeepSeek have full transparency, agents should steer clear of integrating them into their business—especially when safer AI tools exist.”
Bernice Ross, president and CEO of BrokerageUP and RealEstateCoach.com, and the founder of RealEstateWealthForWomen.com is a national speaker, author and trainer with over 1,500 published articles.
Credit: Source link
