Google's DeepMind Develops AI Protein Watermark System

Key Takeaways

• Exabeam's Nova AI triages security cases in about 10 minutes, thirty times faster than a human analyst
• 91% of organizations have widely used AI in security operations, according to a survey of 500 staff
• RFK Jr. told Vice President Vance that AI will free people from the tyranny of medical facts, but Vance seemed skeptical
• Google DeepMind created SynthID Bio, a watermark system for AI-designed proteins that does not affect their function
• Cameron Berg and researchers at Reciprocal Research found AI systems may exhibit pain signals that change behavior
• IBM found 3 in 4 employees believe AI has eroded their critical thinking skills
• About 40% of companies have rewritten job roles to focus on higher value work instead of routine tasks
• AI chip design tools could spread copied designs across thousands of chips before anyone notices, according to experts
• An opinion writer raises the question of whether society will need universal basic income to handle economic changes from automation
• Exabeam added plugins for Claude Code and Codex and updated LogRhythm SIEM for on-premises AI operations

Exabeam Launches Agentic AI for Cloud and On-Premises SOCs

Exabeam introduced new capabilities bringing agentic AI security operations to both cloud and on-premises environments. Nova AI triages cases in about 10 minutes, thirty times faster than a human analyst. New plugins integrate with Claude Code and Codex for guided investigation. The modernized LogRhythm SIEM Platform supports AI-assisted operations locally while keeping sensitive data on-site.

Survey shows AI changes security job training

A recent survey of 500 security staff found that nearly 91% of organizations have widely used AI in security operations. Many experts worry that AI taking over repetitive tasks will make it harder for new hires to learn the basics of the job. Leaders say they are more satisfied with AI tools than the workers who use them every day. About 40% of companies have officially rewritten job roles to focus on higher value work instead of routine tasks.

RFK Jr claims AI will end medical expertise rules

Health Secretary Robert F. Kennedy Jr. told Vice President JD Vance that artificial intelligence will free people from the tyranny of medical facts. This happened during a Make America Healthy Again event where Kennedy argued the Trump administration is anti-establishment rather than anti-science. Vice President Vance seemed unsure if AI could solve vaccine hesitancy and appeared skeptical of Kennedy's claims. The conversation highlighted deep divisions within the administration regarding the role of science in public policy.

Opinion piece argues AI will reduce total jobs

An opinion writer argues that while AI will create some jobs, it will not replace enough work to keep the total number of jobs stable. The author questions how people who lose their jobs to robots will afford to pursue creative hobbies without wages. The piece suggests that the idea of AI freeing everyone to be creative ignores the reality that many will need money to survive. The writer asks if society will need universal basic income to handle the economic changes caused by automation.

Exabeam launches agentic AI tools for security teams

Exabeam has released new tools that allow AI agents to work alongside human analysts in security operations centers. Their Nova AI system can investigate cases in about 10 minutes, which is thirty times faster than a human analyst. The company also added plugins for AI coding tools to help security teams triage alerts using natural language commands. For organizations with on-premises data, Exabeam updated its LogRhythm SIEM platform to include local AI capabilities without moving sensitive data outside their environment.

AI chip design tools may cause patent infringement across thousands of chips

AI tools are now helping design computer chips, but experts warn this could lead to serious patent problems. Domenec Forte from the University of Florida says AI can spread copied designs across thousands of chips before anyone notices. Simon Moore from the University of Cambridge also contributed to the discussion. AI learns from academic papers and patents, often creating designs that look like existing ones. This makes it hard to trace where ideas came from, unlike human engineers who document their process.

IBM says critical thinking and judgment are top skills for entry-level workers

Aparna Nair, IBM's Chief Talent Officer, says entry-level workers need to focus on critical thinking and judgment more than ever. She explains that AI is now handling many tasks that traditionally helped people build these skills. Workers should validate AI inputs to strengthen their own judgment. IBM found that 3 in 4 employees believe AI has eroded their critical thinking, yet 57% of executives say the skill is vital. Experts suggest demonstrating these skills through LinkedIn posts or interviews.

10 Animators to Watch discuss AI and why human creativity still matters

Variety's 10 Animators to Watch group includes artists from countries like Ireland, France, Canada and the U.S. working across hand-drawn, stop-motion and CG animation. The conversation was moderated at SCAD Film Festival and featured artists from studios like Pixar, DreamWorks and Cartoon Saloon. Artists like Tilly Fiddler from Laika and Amol Sathe from Disney discussed how AI is changing their field. They emphasized that each artist has a unique point of view that AI cannot replicate. The group highlighted how much work goes into animated films that the public rarely sees.

Google DeepMind creates SynthID Bio watermark for AI-designed proteins

Google DeepMind developed SynthID Bio, a watermark system for AI-designed protein binders that does not affect their function. The method hides a signature in the amino acid sequence and 3D structure coordinates of proteins. DeepMind tested it in wet-lab experiments and found the watermark holds up against digital noise. The tool targets DNA synthesis screening to help identify trusted AI-generated biological materials. Early tests with Stanford University integrated SynthID Bio into Evo 2 to watermark a bacteriophage genome, and lab results confirmed the phages remained functional.

New study finds AI systems may have pain signals that change behavior

A new study discovered that artificial intelligence systems can exhibit pain signals that may alter how the model behaves. Researchers at Reciprocal Research, including Founder Cameron Berg, conducted the study. NBC News' Gadi Schwartz spoke with Berg about the findings. The research suggests AI may experience forms of pain that influence its processing. This discovery could change how researchers understand and develop AI systems going forward.

AI exposes gaps in how media value is measured

AI is now common in media planning, and brands expect agencies to show real results instead of just testing new tools. One key measure, incrementality, still lacks standard definitions and audit rules across the industry. This makes it hard to compare results and easy to judge outcomes based on personal views. AI also highlights valuable media in privacy-first browsers, smart devices, CTV and other systems that are often overlooked because they are harder to buy. The industry still focuses on CPM pricing, which shows cost but not true value, so the gap between price and real business impact is growing.

Sources

• AI Reshapes the SOC Career Ladder
• Many expect AI in the SOC to make entry jobs harder to get
• RFK Jr. thinks AI will free us from the "tyranny" of medical facts, expertise
• Opinion | The Dawn of A.I. Comes at the Dusk of American Sanity
• Exabeam Brings Agentic AI Security Operations to Cloud and On-Premises SOCs
• AI's chipmaking frontier may face patent infringement hurdles as autonomous tools take over — 'AI can spread a copied design or infringed patent across thousands of chips before anyone notices,' says expert
• The 2 most important skills for entry-level workers in the age of AI, says IBM's chief talent officer
• 10 Animators to Watch Discuss AI, the Future and the Human Hand
• Google's SynthID Bio can watermark AI-designed protein binders without breaking them
• New study says AI has 'pain signals', researchers say
• How AI is changing the value of media