RESEARCH TRIANGLE PARK, N.C., Feb. 25, 2025 (GLOBE NEWSWIRE) -- Data443 Risk Mitigation, Inc. (OTCPK: ATDS) (“Data443” or the “Company”), an AI data security and privacy software company for "All Things Data Security," today announced new capabilitis for its data classificaiton and governance product line - ClassiForAI (CAFAI). This offering leverages the companies' significant and ongoing investments in Machine Learning to accelerate customer adoption of AI & LLMs.
The product capabilty includes access to their new datacenter facilitities in the heart of Research Triangle Park and Data LLM Training engineering. The capability enables customers to come with their own AI engine of choice (Amazon, Microsoft, HuggingFace, ChatGPT, etc.) and have their internal corporate documents form the foundation for a very accurate, safe and confidential AI model for end users, advanced analytics, and of course - to train new AIs. Most importantly, Data443's capability includes identification of extremely sensitive content that is not allowed to be generally exposed in any way by the AIs.
"As we spoke to analysts about our approach, the result was the same - 'Customers are struggling with what do to, don't trust the public cloud, and really do not have the capabilities inside their own business. Plus, they are nervous about disclosure of sensitive content to employees and customers. Customers have no way of separating this data.", stated Jason Remillard, CEO and Founder of Data443.
The methology that Data443 applies is simplistic and focussed on fast results and high accuracy. A known issue with AI engines is that if you train it on too much data, mixed use data or data that is not specific enough - you end up with unreliable models which are prone to problems of hallucianation and unsourced content. Data443's ClassiForAI utilizes its existing capabiltiy of classifying content with over 1,400 policies in 43 languages. The company can scan a massive content repository (of almost any kind in legal, finiance, defence, government) and produce reference examples of extremely high confidence datasets that match the policy - for example - (Personal Privacy Information) and language (German).
'We've been offering our classification engine for different use cases for years, and the feedback is always the same - your policy frameworks are unique and on target. By leveraging our ecosystem (including physical hosting of the models) Data443 is able to provide full lifecycle services for AI accuracy, and reverse train negative outputs for usage in security and disclosure environments. To truely garner the benefits of AI, it isn't useful if it is inaccurate, making up informtation, or its capabilities degrade over time. Our solutions are designed to be a full life cycle implementation - with continues subscriptions in place to continuously refine models, execute data transactions with them and in some cases, host the hardware and softare components on behalf of the customers".
The acquisition coincides with significant market validation of AI-powered email security solutions, evidenced by Abnormal Security's anticipated IPO and growing enterprise demand for intelligent security platforms like Sailpoint. This strategic move positions Data443 to capture an expanding share of the email security market, which is experiencing rapid growth driven by the increasing sophistication of cyber threats and its recent acquisitions of Cyren.
The announcement today will deliver immediate benefits to Data443's customers:
Offline and live training of AI LLMs
Rental of Data443's AI hardware, including NVidia, Tenstorrent, AMD, Cerebras Systems.
Secured facilites in its new USA-based data center.
High power draw capabilities for certain physical premises
Continuous leasing of AI engines for continuous data analysis while it is being used for training or queries.
Identification and removal of extremely sensitive content as defined by the customer.
Reducing exposure of content by LLM's in chatbots, emails and other distribution types
"This offering has been a long time in coming as the industry continues to iterate. Much like our investments with Ripple XRP, these long term plays differentiates us from others as we have mature technology, usually with 1-2 decades of runtime, with actual customers. The startup space has much excitement and investment dollars - which we appreciate. We like our position as recognized experts in data center management and classification," added Remillard. "Like our recent acquisition of Breezemail.ai - we will continue to share with the industry as we win customer engagements."
Data443 Risk Mitigation, Inc. (OTCPK: ATDS) provides software and services to enable secure data across devices and databases, at rest and in flight/in transit, locally, on a network or in the cloud. We are All Things Data Security™. With over 10,000 customers in over 100 countries, Data443 provides a modern approach to data governance and security by identifying and protecting all sensitive data regardless of location, platform or format. Data443’s framework helps customers prioritize risk, identify security gaps and implement effective data protection and privacy management strategies.
Forward-Looking Statements
This press release contains forward-looking statements. These statements are made under the "safe harbor" provisions of the U.S. Private Securities Litigation Reform Act of 1995. Forward-looking statements may be identified by use of terms such as "expect," "believe," "anticipate," "may," "could," "will," "should," "plan," "project," "intend," "estimate," "predict," "potential," "pursuant," "target," "continue" or the negative of these words or other comparable terminology. Statements in this press release that are not historical statements, including statements regarding Data443's plans, objectives, future opportunities for Data443's services, future financial performance and operating results, and any other statements regarding Data443's future expectations, beliefs, plans, objectives, financial conditions, assumptions or future events or performance, or regarding the anticipated consummation of any transaction, are forward-looking statements. These statements are not guarantees of future performance and are subject to numerous risks, uncertainties and assumptions, many of which are difficult to predict or are beyond Data443's control. These risks, uncertainties and assumptions could cause actual results to differ materially from the results expressed or implied by the statements. They may relate to the outcome of litigation, settlements and investigations; actions by third parties, including governmental agencies; volatility in customer spending; global economic conditions; inability to hire and retain personnel; loss of, or reduction in business with, key customers; difficulty with growth and integration of acquisitions; product liability; cybersecurity risk; anti-takeover measures in the Company’s charter documents; and the uncertainties created by global health issues, such as the ongoing outbreak of COVID, and political unrest and conflict, such as the invasion of Ukraine by Russia. These and other important risk factors are described more fully in the Company’s reports and other documents filed with the Securities and Exchange Commission ("the SEC"), including in Part I, Item 1A of the Company’s Annual Report on Form 10-K filed with the SEC on April 17, 2024, and subsequent filings with the SEC. Undue reliance should not be placed on the forward-looking statements in this press release, which are based on information available to the Company on the date hereof. Except as otherwise required by applicable law, Data443 undertakes no obligation to publicly update or revise any forward-looking statements, whether because of new information, future events or otherwise.
“DATA443” is a registered trademark of Data443 Risk Mitigation, Inc.
All product names, trademarks and registered trademarks are property of their respective owners. All company, product and service names used in this press release are for identification purposes only. Use of these names, trademarks and brands does not imply endorsement.
