- Smarter with AI
- Posts
- SunBrief#15: Deepseek’s Data Breach Exposed
SunBrief#15: Deepseek’s Data Breach Exposed
EU bans unacceptable AI
In partnership with
Welcome to the SunBrief
Today in SunBrief 🌞
AI free guide to Master ChatGPT
Deepseek’s Data Breach Exposed
Stock Updates
Master AI in 5 Minutes a Day
OpenAI’s Key Partnerships and AI Advancements
AI Highlights of the Week
EU bans “Unacceptable Risk" AI
Too Important to Miss
AI free guide to Master ChatGPT
Want to get the most out of ChatGPT?
ChatGPT is a superpower if you know how to use it correctly.
Discover how HubSpot's guide to AI can elevate both your productivity and creativity to get more things done.
Learn to automate tasks, enhance decision-making, and foster innovation with the power of AI.
Deepseek’s Data Breach Exposed
Exposed DeepSeek Database Revealed Chat Prompts and Internal Data
A security lapse at DeepSeek has exposed sensitive internal data, including user chat prompts and proprietary information, raising fresh concerns about the Chinese AI company’s data practices.
Here’s what’s happening:
Database Leak: An unsecured DeepSeek database was discovered online, revealing user interactions, internal company documents, and technical logs. The leaked data included unfiltered chat prompts, exposing how users interacted with the AI.
Security and Privacy Risks: The breach highlights vulnerabilities in DeepSeek’s data protection measures, amplifying existing concerns about the safety of using its AI tools, particularly given its ties to China.
Impact on Trust: This incident follows recent scrutiny over DeepSeek’s alleged use of OpenAI’s models and the US Navy’s ban on its apps, further damaging its credibility in global markets.
Why It Matters:
The leak underscores the risks of rapid AI development without robust security safeguards. For DeepSeek, already under fire for alleged intellectual property issues and ethical concerns, this breach adds fuel to debates about the reliability and transparency of its technology.
Does the DeepSeek database leak affect your trust in its AI tools? |
Stock Updates
Master AI in 5 Minutes a Day
Learn AI in 5 minutes a day
What’s the secret to staying ahead of the curve in the world of AI? Information. Luckily, you can join 1,000,000+ early adopters reading The Rundown AI — the free newsletter that makes you smarter on AI with just a 5-minute read per day.
OpenAI’s Key Partnerships and AI Advancements
OpenAI Expands AI Capabilities with New Partnerships and Models
OpenAI is advancing AI research with key initiatives, including a partnership with U.S. National Laboratories, the launch of o3-mini, and the introduction of ‘deep research’ for autonomous analysis.
Here’s what’s happening:
AI Collaboration with U.S. National Labs: OpenAI is working with the U.S. National Laboratories to enhance scientific research using AI models. This partnership will accelerate breakthroughs in energy, materials science, and physics by deploying AI on supercomputers.
Launch of OpenAI o3-mini: The o3-mini model is designed for cost-efficient reasoning in STEM fields, offering improved coding, math, and scientific problem-solving with lower latency. It is now available via ChatGPT and API.
Introducing 'Deep Research': OpenAI unveiled ‘deep research,’ an AI tool for autonomous multi-step research tasks. Leveraging the o3 model, it synthesizes online information into comprehensive reports, assisting professionals with data-driven insights.
These efforts reinforce OpenAI’s commitment to driving AI innovation, advancing research, and making AI-powered tools more accessible.
Which OpenAI initiative excites you the most? |
AI Highlights of the Week
Google’s 'Ask for Me' AI Feature to Call Businesses on your behalf?
Google is piloting "Ask for Me," an AI-driven tool that makes phone calls to local businesses on users' behalf to inquire about services and pricing. Currently limited to nail salons and auto repair shops, users can opt into the experiment via Google Search Labs. The AI introduces itself as an automated system from Google during calls, and businesses can choose to opt out of receiving such calls. This feature builds upon Google's Duplex technology.
OpenAI’s o1 reasoning model free for all Copilot users
Chinese AI startup MiniMax has released the MiniMax-01 series, featuring MiniMax-Text-01 and MiniMax-VL-01 models. MiniMax-Text-01, a language model with 456 billion parameters, can process up to 4 million tokens, enabling long-context applications. These open-source models position MiniMax as a competitor to leading AI systems.
Google's Gemini 2.0 Flash is now ravailable for everyone
Google has updated its Gemini app to incorporate the Gemini 2.0 Flash model, offering users faster responses and improved performance for tasks like brainstorming, learning, and writing. This enhancement is now available on both web and mobile platforms.
Chinese State-Linked Accounts Promoted DeepSeek AI Before U.S. Stock Decline
Chinese state-linked social media accounts, including those of diplomats and state media, actively promoted DeepSeek's AI model launch, highlighting China's advancements in AI. This campaign preceded a significant downturn in U.S. tech stocks.
NHS Launches Largest AI Breast Cancer Screening Trial
The NHS is initiating the world's largest trial of AI in breast cancer screening, involving around 700,000 mammograms across England. The study will assess AI's effectiveness compared to radiologists in detecting breast cancer. If successful, AI could serve as a second reader, potentially reducing the need for dual radiologist reviews and expediting diagnoses. The £11 million trial is funded by the National Institute for Health and Care Research and is expected to conclude in several years.
EU bans “Unacceptable Risk" AI
EU is Now Enforcing the AI Act, Banning "Unacceptable Risk" AI Systems
As of February 2, 2025, the European Union has enacted a ban on artificial intelligence systems classified as posing an "unacceptable risk." This move is part of the EU's comprehensive Artificial Intelligence Act, which aims to regulate AI applications based on their potential impact on safety and fundamental rights.
Key Provisions:
Prohibited AI Practices: The Act explicitly bans AI systems that manipulate human behavior, exploit vulnerabilities of specific groups, or employ real-time biometric identification in public spaces without proper authorization. Additionally, AI applications used for social scoring by governments, which assess individuals based on personal characteristics or behaviors, are prohibited.
High-Risk AI Applications: AI systems utilized in critical sectors such as healthcare, education, employment, and law enforcement are designated as high-risk. These applications must adhere to stringent requirements, including rigorous risk assessments, data governance protocols, and human oversight mechanisms, to ensure their safe deployment.
Enforcement and Penalties: Organizations found in violation of these regulations may face substantial fines, potentially reaching up to 7% of their global turnover or €35 million, whichever is greater. This stringent enforcement underscores the EU's commitment to responsible AI development and deployment.
The implementation of these measures reflects the EU's proactive stance in establishing a regulatory framework that balances innovation with ethical considerations, aiming to foster public trust in artificial intelligence technologies.
Do you support the EU’s decision to ban AI systems with ‘unacceptable risk’? |
Too Important to Miss
Last Week’s Poll Result
Do you think DeepSeek will overthrow ChatGPT/OpenAI as the leading AI model?
Yes, 35.09%. Maybe, 33.33%. No, 31.58%.
Do you think DeepSeek’s success is built on OpenAI’s work?
Yes, 56.86%. No, 13.73%. Unclear, 29.41%.
Which AI model do you think governments should adopt for public sector use?
ChatGPT-Gov (OpenAI), 50.00%. LLaMA AI (Meta), 6.25% DeepSeek’s Janus-Pro-7B, 31.25%. Others, 12.5%.
Feedback
We’d love to hear from you!How did you feel about today's SunBrief? Your feedback helps us improve and deliver the best possible content. |
Know someone who may be interested?
And that's a wrap on today’s SunBrief!
Reply