Hive, a number one supplier of enterprise AI options, has been awarded a Division of Protection (DoD) contract for deepfake detection of video, picture, and audio content material. This groundbreaking partnership marks a decisive leap ahead in safeguarding nationwide safety and combatting the escalating dangers of artificial media and AI-generated disinformation.
Additionally Learn: AiThority Interview with Joe Fernandes, VP and GM, AI Enterprise Unit at Purple Hat
“As frontline defenders towards AI-generated deception, we perceive what’s at stake and are dedicated to delivering unmatched detection accuracy and pace, empowering U.S. protection forces with cutting-edge instruments that keep forward of potential threats.”
Beneath the preliminary two-year contract, Hive will work with the Protection Innovation Unit (DIU) to help the intelligence group with Hive’s state-of-the-art deepfake detection fashions, deployed in an offline, on-premise surroundings and able to detecting AI-generated video, picture, and audio content material. With AI manipulation now a weapon of alternative for adversarial entities, Hive’s progressive know-how stands as a digital safeguard, defending the integrity of significant info ecosystems and reinforcing the DoD’s preparedness for future unfold of misinformation, threats, and conflicts.
“Hive is honored to hitch forces with the Division of Protection on this essential mission,” stated Kevin Guo, CEO and Co-Founding father of Hive. “As frontline defenders towards AI-generated deception, we perceive what’s at stake and are dedicated to delivering unmatched detection accuracy and pace, empowering U.S. protection forces with cutting-edge instruments that keep forward of potential threats.”
Additionally Learn: OpenPad AI Raises $2Million To Remodel Web3 Investments With Decentralized AI Analytics
Since 2017, Hive has emerged because the main know-how firm utilizing AI to determine dangerous on-line content material, analyzing billions of items of content material every month from tons of of main corporations throughout industries together with Reddit, Zynga, Midjourney, Runway, BlueSky, Reality Social, and extra. Along with the identification of deepfakes and AI-generated content material, Hive’s capabilities embrace detection of a complete set of specific and dangerous content material varieties, together with little one sexual abuse materials (CSAM) in partnership with little one security chief, Thorn. Hive’s fashions can be found via a number of deployment choices, together with cloud-based APIs in addition to on-premise and personal cloud deployments via a partnership with NVIDIA NIM.
As threats proceed to emerge and escalate, Hive’s dedication stays unwavering: to offer the world’s most succesful AI fashions for validating the protection and authenticity of digital content material.
[To share your insights with us as part of editorial or sponsored content, please write to psen@itechseries.com]