Artificial Intelligence | News | Insights | AiThority
[bsfp-cryptocurrency style=”widget-18″ align=”marquee” columns=”6″ coins=”selected” coins-count=”6″ coins-selected=”BTC,ETH,XRP,LTC,EOS,ADA,XLM,NEO,LTC,EOS,XEM,DASH,USDT,BNB,QTUM,XVG,ONT,ZEC,STEEM” currency=”USD” title=”Cryptocurrency Widget” show_title=”0″ icon=”” scheme=”light” bs-show-desktop=”1″ bs-show-tablet=”1″ bs-show-phone=”1″ custom-css-class=”” custom-id=”” css=”.vc_custom_1523079266073{margin-bottom: 0px !important;padding-top: 0px !important;padding-bottom: 0px !important;}”]

IronCore Labs Launches Beta for Encrypting AI Data

In response to mounting risks facing software companies with AI solutions, IronCore Labs announced a beta signup today for a first-of-its-kind product called Cloaked AI.

Cloaked AI is an encryption-in-use solution that allows the AI memory – the vector embeddings – to be used fully without loss of functionality while also protecting the privacy and security of that data no matter who holds it. Only those with the key can meaningfully query the data and derive meaning from it.

AiThority: How AI Can Improve Public Safety

Companies with AI tools store personal, sensitive information in unprotected databases
It’s fairly well known that a handful of AI models like OpenAI and ChatGPT now power thousands of AI products built by software companies across the globe. What’s less well known is that advanced AI features require long-term memory – a place to store context, history, customer information, and other remembered impressions of data. This information then gets stored in a vector database.

Vector databases now hold the crown jewels
AI products have access to all kinds of private data, from chat history to medical diagnoses to proprietary company information, which is represented by vector embeddings and stored in these new vector databases. While these vector embeddings aren’t understandable to humans, much like other machine representations of data, they hold a ton of meaning for AI and a savvy hacker can easily extract highly sensitive information from them for mischief or profit.

An encryption solution that won’t disrupt your business, but will disrupt a hacker
Protecting AI data stored in vector databases is critical to securing AI systems, ensuring trust with customers or vector database vendors, complying with global data privacy regulations and avoiding privacy and security breaches.

Related Posts
1 of 40,907

Read More: The Practical Applications of AI in Workplace

Thanks to the advanced encryption-in-use technologies at play in Cloaked AI, all major operations like nearest neighbor searches, clustering, anomaly detection, and more, work over all known vector databases including those from Pinecone, Weaviate, Elastic, and AWS OpenSearch.

Cloaked AI is built to be invisible to you but impenetrable to attackers. You continue business as usual. Your customers get the privacy protection they want. Hackers get nothing.

Sign up to be the first to try Cloaked AI
Cloaked AI allows companies to benefit from the power of LLMs, generative AI, semantic search, and related breakthroughs that require embeddings while protecting private data.

Latest Insights: What Techniques Will Deliver for Measuring Attention in 2023?

[To share your insights with us, please write to sghosh@martechseries.com]

Comments are closed.