April 29, 2025

news   Read News
DeepSeek R2, Protect AI, HP, Aqua Security, AI Security, SABER Project

The AI landscape is rapidly evolving, with several significant developments shaping the industry. DeepSeek is set to release its new AI model, DeepSeek R2, which promises to be more efficient and powerful than its predecessor, potentially disrupting the market. Meanwhile, Protect AI is leading conversations on GenAI and AI security at the RSAC Conference 2025, highlighting the importance of securing AI applications. The US military is also seeking to assess AI vulnerabilities through the Securing Artificial Intelligence for Battlefield Effective Robustness (SABER) project. On the commercial front, HP has introduced new AI-integrated products, and Aqua Security has launched a comprehensive security solution for AI-powered applications. However, Americans are skeptical about the impact of AI on the news industry, with many expressing concerns about job losses and inaccurate information. The White House is seeking public input to revise the national plan for AI research and development, while the Copyright Licensing Agency is developing a collective licensing solution for GenAI training. AI is also being explored for its potential to mitigate the shortage of radiologists and improve care quality. Overall, the AI industry is poised for significant growth and transformation, with a focus on security, ethics, and responsible development.

Key Takeaways

  • DeepSeek's new AI model, DeepSeek R2, is expected to be more efficient and powerful than its predecessor, with 1.2 trillion parameters and a rumored cost of $0.07 per million input tokens.
  • Protect AI is leading conversations on GenAI and AI security at the RSAC Conference 2025, emphasizing the importance of securing AI applications.
  • The US military is seeking to assess AI vulnerabilities through the Securing Artificial Intelligence for Battlefield Effective Robustness (SABER) project.
  • HP has introduced new AI-integrated products, including the OmniBook Ultra Flip 14 and the HP EliteBook X G1a.
  • Aqua Security has launched a comprehensive security solution for AI-powered applications, providing governance and policy-driven detection and response.
  • Americans are skeptical about the impact of AI on the news industry, with concerns about job losses and inaccurate information.
  • The White House is seeking public input to revise the national plan for AI research and development.
  • The Copyright Licensing Agency is developing a collective licensing solution for GenAI training, ensuring compensation and remuneration for publishers and authors.
  • AI has the potential to mitigate the shortage of radiologists and improve care quality, but its impact on the radiology profession and patient care needs further investigation.
  • Broadcom's stock price has fallen 2% due to rumors of DeepSeek's upcoming AI model, highlighting the potential market impact of AI developments.

DeepSeek R2 AI Model Coming Soon

DeepSeek is set to release its new AI model, DeepSeek R2, which is expected to be more efficient and powerful than its predecessor. The model will use Huawei Ascend 910B chips and have 1.2 trillion parameters. DeepSeek R2 is rumored to be 97.3% cheaper to train than GPT-4 and will cost around $0.07 per million input tokens. The release of DeepSeek R2 may impact the US stock market and the AI industry as a whole. DeepSeek's previous model, R1, caused a stir in the market due to its low cost and high performance. The company's use of open-source software and local supply chain may also affect the industry.

Broadcom Stock Falls Due to DeepSeek AI

Broadcom's stock price has fallen due to rumors of DeepSeek's upcoming AI model, DeepSeek R2. The new model is expected to be more efficient and cost-effective than its predecessor and may compete with OpenAI's o3 and o4-mini. DeepSeek's previous model, R1, was released at a fraction of the cost of similar models and caused a stir in the market. The company's use of Huawei Ascend 910B chips may also affect the market. Broadcom's stock price has fallen 2% due to the rumors, and the company's future performance may be impacted by the release of DeepSeek R2.

Protect AI Leads AI Security Conversations

Protect AI is participating in the RSAC Conference 2025 with a focus on securing GenAI and AI supply chains. The company's Chief Information Security Officer, Diana Kelley, and other experts will deliver four official sessions on AI security. Protect AI will also host fireside chats with AI security leaders and practitioners. The company has been named a Visionary in AI Security and is recognized for its leadership in addressing emerging AI threats. Protect AI's participation in the conference aims to promote safe innovation in AI and provide solutions for securing AI applications.

Protect AI at RSAC Conference 2025

Protect AI is leading conversations on GenAI and AI security at the RSAC Conference 2025. The company's CISO, Diana Kelley, and other experts will present four official sessions on AI security and host fireside chats with industry leaders. Protect AI has been recognized for its leadership in AI security and aims to promote safe innovation in the field. The company's participation in the conference highlights the importance of securing AI applications and addressing emerging threats.

AI Helps with Radiologist Shortage

Artificial intelligence (AI) can help mitigate the shortage of radiologists by interpreting images and reducing workload. A new review of literature suggests that AI can assist radiologists in high and low-likelihood cases, reducing the number of exams to be conducted. However, ethical, regulatory, and legal issues may challenge the deployment of AI in radiology. Researchers recommend that trainees not rely solely on AI and that organizations ensure radiologists have the final decision in a transition period. AI can improve care quality and resilience, but its impact on the radiology profession and patient care needs further investigation.

US Military Seeks to Assess AI Vulnerabilities

The US Defense Advanced Research Projects Agency (DARPA) is extending the deadline for a project to assess the vulnerabilities of military AI systems to cyber attacks. The Securing Artificial Intelligence for Battlefield Effective Robustness (SABER) project aims to build an AI research group to evaluate AI-enabled battlefield systems. The project seeks to develop techniques, tools, and technical competency to assess AI vulnerabilities and ensure the security of AI-enabled systems. The deadline for proposals has been extended to June 3, 2025.

HP Introduces New AI-Integrated Products

HP has introduced a new line of next-gen AI-integrated products, including the OmniBook Ultra Flip 14 and the HP EliteBook X G1a. The products feature AI-enhanced evolution, robust protection, and seamless collaboration. HP also introduced AI software experiences, including HP AI Companion and Poly Camera Pro. The company aims to provide practical solutions for the future of work by prioritizing AI across all product lines. The Vietnamese government's AI Readiness Index has risen, and HP's 2024 Work Relationship Index reports that only 28% of workers have a healthy relationship with work.

Aqua Security Introduces Secure AI

Aqua Security has introduced Secure AI, a comprehensive security solution designed to safeguard AI-powered applications throughout their lifecycle. The solution provides governance and policy-driven detection and response, ensuring AI applications maintain integrity without requiring modifications to existing code or workflows. Aqua Security aims to protect the evolving landscape of applications via deep runtime protection and offers features such as AI code scanning and runtime detection of AI threats.

Americans Skeptical of AI Impact on News

A Pew Research Center survey found that Americans are pessimistic about the long-term effect of artificial intelligence on the news people get and the journalism profession. Roughly half of US adults say AI will have a negative impact on the news, while 10% say it will have a positive effect. The survey also found that 59% of Americans say AI will lead to fewer jobs for journalists, and 66% are concerned about people getting inaccurate information from AI. The results suggest that Americans are skeptical of AI's impact on the news industry and journalism.

Collective Copyright Licence for GenAI

The Copyright Licensing Agency (CLA) is developing a Generative AI (GenAI) Training Licence, which aims to provide a scalable collective licensing solution for GenAI training. The licence will ensure compensation and remuneration for publishers and authors while providing developers with legal certainty and material to train models. The CLA is working with the Publishers' Licensing Services (PLS) and the Authors' Licensing and Collecting Society (ALCS) to develop the licence, which is expected to be published in the third quarter of 2025.

White House Seeks Input on AI Research Plan

The White House is seeking public input to revise the national plan for artificial intelligence research and development. The Office of Science and Technology Policy has released a request for information to gather comments on how to maintain the country's competitiveness in AI. The current plan was updated in 2023, and the White House is looking to revise it to reflect the latest developments in AI. The comment period will last until May 29, and the revised plan will aim to ensure the US remains a leader in AI research and development.

Sources

AI DeepSeek R2 Artificial Intelligence GenAI AI Security AI Vulnerabilities