F5 is placing AI security and infrastructure resilience at the center of its business strategy. Jason Baden, F5 regional vice president, said companies are rethinking where AI workloads run, with many moving workloads back on-premises to maintain data control. Security discussions have moved into the boardroom, and patching timelines are shrinking from weeks to hours. AI security and maximizing existing infrastructure are expected to be the two biggest growth areas over the next 12 months.
New IDC research sponsored by Portnox highlights a growing security concern: AI agent access. For 2026, agent-to-agent access ranked first at 18%, followed by agent-to-resource at 16% and agent-to-device at 15%. Because AI agents can act without human oversight, organizations are looking for faster ways to block risky access and track how quickly threats are contained after detection.
President Trump signed an executive order rebranding AI as Super Intelligence, or SI, arguing the term better reflects how the technology amplifies human ability. The term is appearing in some government documents, but adoption remains mixed. Many officials and companies still use AI, and Trump has publicly pushed for the new term to become standard, even criticizing outlets that did not adopt it.
Researchers from USC, MIT, UC Berkeley and the Psychology of Technology Institute launched ImpactBench, an open platform that evaluates AI models across 800 metrics and 26 benchmarks. The framework measures physical, psychological and social effects and converts the results into AI Nutrition Labels. The goal is to help people understand how AI systems affect human well-being, beyond just speed and accuracy.
KPMG says heavy spending on data infrastructure has not solved data quality problems for AI. The challenge is not finding data but understanding its history, context and meaning. KPMG identifies searchability, context and trustworthiness as three requirements for AI-ready data. Experts say AI now makes it easier to justify the business value of improving data governance.
Robotics is advancing through improvements in both hardware and software. New vision-language-action models allow robots to follow plain language instructions and handle tasks they were not specifically trained for. Amazon's Vulcan robotic arm manages about 75 percent of over one million catalogue items. Humanoid robots remain limited to repetitive tasks in structured settings, with household robots likely more than 20 years away.
Netflix introduced an AI Discover tab that lets users find shows and movies based on mood. Users can select themes like something light or scare me and receive personalized recommendations. The feature is rolling out in the U.S. to English-speaking TV users first, and a Tune Your Recommendations button lets users save or exclude suggestions.
At Wayzata High School, students used Gemini Canvas to build apps that tracked athletic performance during a two-day scouting combine. Teacher Riley Davis designed the project to teach vibe coding, where AI helps create software through written instructions. Students practiced prompt engineering and adjusted their instructions when results were unexpected, covering basketball shooting, soccer drills and the 40-yard dash.
A new study using complex algorithms to simulate diagnostic reasoning found that AI can outperform physicians on diagnostic tests. The research moves beyond simple medical quizzes to mimic the back-and-forth of real diagnosis. The study highlights the gap between medical knowledge and clinical judgment and points to potential implications for future diagnostic tools.
OpenAI's Caroline Zier urged the US to stay alert during upcoming AI talks with China. She warned about alleged Chinese distillation attacks that could threaten US technology. Zier supported dialogue with Beijing but said she doubts both sides share the same understanding of AI risks. She previously worked for the US Department of Defence.
Key Takeaways
- F5 regional vice president Jason Baden said AI security and infrastructure optimization are the two biggest growth areas for the next 12 months.
- IDC research shows AI agent-to-agent access is the top security priority for 2026 at 18%, followed by agent-to-resource at 16%.
- Trump signed an executive order rebranding AI as Super Intelligence (SI), though adoption of the term remains mixed across government and industry.
- ImpactBench evaluates AI models across 800 metrics and 26 benchmarks, producing AI Nutrition Labels focused on human well-being.
- KPMG identifies searchability, context and trustworthiness as the three requirements for AI-ready data.
- Amazon's Vulcan robotic arm handles about 75% of over one million catalogue items using vision-language-action models.
- Netflix launched an AI Discover tab for mood-based recommendations, rolling out first to English-speaking US TV users.
- Wayzata High School students used Gemini Canvas to build athletic performance tracking apps through vibe coding.
- A diagnostic simulation study found AI can outperform physicians on diagnostic tests that mimic real-world reasoning.
- OpenAI warns the US to stay alert in China AI talks, citing alleged Chinese distillation attacks on technology.
F5 shifts focus to AI security and infrastructure resilience
F5 is prioritizing AI security, application security, and digital resilience as companies adopt new technologies faster. Jason Baden, F5 regional vice president, said businesses now care about data sovereignty and securing AI environments. Security talks have moved to boardrooms where leaders want clear patching timelines and real-time protection. F5 is expanding its product lineup and partnering with channel providers to meet growing demand.
F5 says AI and sovereignty are reshaping infrastructure
F5 regional vice president Jason Baden said organizations are deciding where AI workloads run and how to keep data under control. Many companies are moving workloads back on-premises instead of relying on the cloud. Security is now a board-level concern, and patching windows are shrinking from weeks to hours. AI security and maximizing existing infrastructure are the two biggest growth areas for the next 12 months.
AI agent access becomes a top security priority for 2026
New IDC research sponsored by Portnox found that AI agent access is a leading security priority for 2026. Agent-to-agent access ranked first at 18%, followed by agent-to-resource at 16% and agent-to-device at 15%. AI agents can act without human oversight, so companies need faster ways to block risky access. The report recommends tracking how quickly threats are contained after detection.
Who is adopting Trump's Super Intelligence term
President Trump signed an executive order rebranding AI as Super Intelligence, or SI, claiming it better reflects how the technology amplifies human ability. The term is slowly appearing in government documents, but many officials and companies still use AI. Trump has pushed for the new term to become standard, even criticizing outlets like the Wall Street Journal. Some tech leaders have started using the term, but adoption remains mixed.
New framework adds AI nutrition labels for well-being
Researchers from USC, MIT, UC Berkeley and the Psychology of Technology Institute created ImpactBench, an open platform that evaluates AI models. It measures 800 metrics across 26 benchmarks covering physical, psychological and social effects. The results are turned into AI Nutrition Labels that show how AI systems affect human well-being. The goal is to help people understand AI's impact beyond speed and accuracy.
Infrastructure spending alone cannot solve AI data trust problems
Organizations have spent heavily on data infrastructure for AI, but data quality remains poor. The main issue is not finding data but understanding its history, context and meaning. KPMG identifies three requirements for AI-ready data: searchability, context and trustworthiness. Experts say AI now makes it easier to justify the business value of improving data governance.
Robotics advances with AI and humanoid designs
Robotics is advancing rapidly through improvements in both hardware and software. New vision-language-action models let robots follow plain language instructions and perform tasks they were not specifically trained for. Amazon's Vulcan robotic arm handles about 75 percent of over one million catalogue items. Humanoid robots are still limited to repetitive tasks in structured settings, with household robots likely more than 20 years away.
Netflix launches AI Discover tab for mood-based watching
Netflix introduced an AI Discover tab that helps users find shows and movies based on their mood. Users can pick themes like something light or scare me and get personalized recommendations. The feature rolls out in the U.S. to English-speaking TV users first. A Tune Your Recommendations button lets users save or exclude suggestions.
Wayzata students use AI to track sports performance
Wayzata High School students used Gemini Canvas to build apps that tracked athletic performance during a two-day scouting combine. Teacher Riley Davis designed the project to teach vibe coding, where AI helps create software through written instructions. Students learned prompt engineering and adjusted their instructions when results were unexpected. Activities included basketball shooting, soccer drills and the 40-yard dash.
AI shows promise in medical diagnosis simulations
A new study used complex algorithms to simulate real-world diagnostic reasoning and found AI can outperform physicians on diagnostic tests. The research moves beyond simple medical quizzes to mimic the back-and-forth of actual diagnosis. The study highlights the gap between medical knowledge and clinical judgment. Results point to important implications for future diagnostic tools.
OpenAI warns US to stay alert in China AI talks
OpenAI's Caroline Zier urged the US to be careful during upcoming AI talks with China. She warned about alleged Chinese distillation attacks that could threaten US technology. Zier supported dialogue with Beijing but doubted both sides share the same understanding of AI risks. She previously worked for the US Department of Defence.
Sources
- F5 evolves strategy to address AI security and infrastructure challenges
- F5 sees AI, security, sovereignty reshaping infrastructure
- AI Agent Access Becomes a Top Security Priority for 2026
- Here’s who is falling in line for Trump’s “Super Intelligence” loyalty test
- What if AI came with a nutrition label? A new framework measures AI's impact on human well-being
- Why infrastructure spending won’t fix the real AI problem
- Executive summary: Robotics comes of age
- Netflix's new AI Discover tab helps you find something to watch based on your mood
- Wayzata Students Combine Technology, Sports in AI Project
- Does AI diagnose better—or take a different test?
- OpenAI executive urges caution in China-US AI talks, warns of distillation
Comments
Please log in to post a comment.