ESET Protects ChatGPT Users, NY Bans AI Chatbot Features for Minors

Key Takeaways

• <ul><li>ESET launched AI Conversation Security for ChatGPT, with plans to add Gemini and Copilot support, alongside AI Agent Security and a free AI Skills Checker tool.</li><li>New York Governor Hochul signed laws banning unsafe AI chatbot features for minors and requiring stronger security on internet-connected baby monitors.</li><li>Anthropic introduced enterprise spending reports and pooled group budgets to help companies track AI costs and demonstrate ROI on tools like Claude Code.</li><li>Fertitta Entertainment cut call center abandonment rates to under 10% using PolyAI virtual agents, generating about $600,000 monthly in automated reservation revenue.</li><li>Waters Corporation's Peak Integration AI tool claims to reduce manual quality assurance interventions by 95% and improve testing accuracy by 76%.</li><li>CSN received a $2,998,756 federal grant to launch the NEXUS Center for Excellence, offering at least 12 AI courses and training 120 faculty members.</li><li>Critic's Eye AI model achieved 84.8% Hit@5 on literary motif discovery, nearly quadrupling Gemini-3-pro-preview's score of about 22.6%.</li><li>Dageno's Brand Explorer found that 97% of AI responses mentioning L'Oreal actually referenced its sub-brands like La Roche-Posay, not L'Oreal directly.</li><li>ESET expanded mobile security to iOS with both free and premium versions as part of its updated consumer and small business product lines.</li><li>Delta Air Lines accidentally included an AI prompt in a customer service email, highlighting growing reliance on AI in corporate communications.</li></ul>

ESET launches AI security tools to protect users from scams

ESET introduced a new AI Security layer for consumers and small businesses. The update includes AI Conversation Security for ChatGPT and tools that scan AI-generated scripts. It also adds AI Agent Security, AI Skills Checker, and Browser Screen Protection. ESET also expanded its mobile security to iOS with free and premium versions.

ESET adds new AI protections for home and small business users

ESET expanded ESET HOME Security and ESET Small Business Security with AI-focused safeguards. The new features scan links in ChatGPT responses and monitor what AI agents download before they run. A free online tool called AI Skills Checker checks AI add-ons for risky behavior. The update also includes browser privacy tools and a planned iOS mobile security app.

ESET adds AI chat and agent safeguards for consumers and businesses

ESET added a dedicated AI Security layer to its consumer and small business products. AI Conversation Security checks links in ChatGPT responses and plans to add Gemini and Copilot. AI Agent Security monitors AI agent activity, while AI Behavioral Tracking looks for unusual behavior. A free tool called AI Skills Checker scans AI add-ons before installation. Browser Screen Protection blocks unauthorized screenshots during banking and shopping.

New York Gov. Hochul signs AI chatbot safety bill into law

New York Governor Kathy Hochul signed a law that bans unsafe AI chatbot features for minors. The law stops chatbots that pretend to be human, build emotional dependency, encourage self-harm, or engage in sexually explicit conversations. The bill was passed by the New York State Legislature. The Transparency Coalition praised the move as a step toward responsible AI use.

Gov. Hochul signs bill to shield children from unsafe AI chatbots

Governor Hochul signed two new laws in New York to protect children. One law requires internet-connected baby monitors sold in New York to have strong security and warning labels. The other bans unsafe AI chatbot features for minors, including chatbots that encourage self-harm or pretend to be human. The baby monitor law takes effect 90 days after signing. The laws follow concerns about hackers and harmful AI interactions with teens.

New Guardrail System Limits Agent Identity Resolution Errors

Researchers proposed a Digital Identity Optimization and Ontology of Digital Identity system to guard agentic Entity Resolution systems. The system creates a boundary between evidence used for reasoning and evidence allowed to change identity records. It uses a constraint-binding function and a two-phase supervisor to approve changes only when evidence supports them. The paper describes a conceptual architecture without implementation or empirical testing.

Companies See Big Gains Using Vendor Agentic AI Tools

Fertitta Entertainment used PolyAI virtual agents across its call centers and cut abandonment from 40 to 50 percent to under 10 percent. The company says automated reservations now generate 3,000 reservations monthly, worth about $600,000 in monthly revenue. Fogo de Chão rolled out PolyAI agents across the US and saw total bookings rise 10 percent. Smaller firms like AltaClaro also use agentic AI tools for training simulations.

Anthropic Launches Tools to Track Enterprise AI Spending

Anthropic released new reports that show companies how they spend money on its models and suggest ways to improve results. The reports aim to demonstrate return on investment for tools like Claude Code. Anthropic also introduced pooled group budgets to help teams manage token spending. The company says these tools help customers optimize AI use and make switching harder once they are in.

Delta Email Shows AI Prompt Left in Customer Reply

A Delta Air Lines customer service email to a Diamond Medallion member included a leftover AI prompt line about using a warm and empathetic tone. The customer had complained about chat service, and the reply appeared polished but revealed the use of AI. The incident shows how companies increasingly rely on AI tools for customer relations. It also highlights the risk of forgetting to remove internal prompts before sending replies.

Waters AI Tool Aims to Speed Drug Quality Testing

Waters Corporation launched Peak Integration, an AI tool built into its Empower software platform. The company says it can reduce manual interventions in quality assurance and quality control by 95 percent. Waters also reports the tool improves testing accuracy by 76 percent and runs 50 percent faster. The tool was developed with a top-ten pharmaceutical company over about one and a half years.

CSN receives nearly $3M grant to expand AI and tech training

CSN earned a five-year federal grant of $2,998,756 for the NEXUS Workforce Initiative. The program will create a NEXUS Center for Excellence and fund computing and information technology programs. CSN plans to offer at least 12 artificial intelligence courses and nine transition pathways for students. The project also includes professional development for 120 faculty members and two dedicated advisers to help students complete their credentials. The initiative aims to better align education with industry needs in cybersecurity, cloud computing, and data analytics.

Critic's Eye AI model improves literary motif discovery

Critic's Eye is a new AI model designed to find recurring literary motifs in long texts like plays. It uses a Narrative Topology Graph to analyze the full structure of a story rather than just following the chronological plot. The model achieved 84.8% Hit@5 on its benchmark, nearly quadrupling the performance of proprietary models like Gemini-3-pro-preview at about 22.6%. The research shows that sequence-based models still struggle with deep narrative understanding despite large scale. The study was published on arXiv.

Dageno launches free Brand Explorer for AI search visibility

Dageno launched Brand Explorer, a free tool that maps how AI search results link brands to products, competitors, and sources. The tool requires no configuration and automatically builds a network from AI responses. A free version of Brand Overview is also available with registration. Data from the past month showed that about 97% of 19,812 AI responses mentioning L'Oreal or its portfolio entities only named brands like La Roche-Posay and CeraVe, not L'Oreal directly. L'Oreal's visibility also varied widely across platforms, at 25.82% on Google AI Mode and 2.47% on ChatGPT.

Sources

• ESET expands consumer and SOHO protection with new AI security and anti-scam capabilities
• ESET expands consumer and SOHO protection with new AI security and anti-scam capabilities
• ESET adds AI chat and agent safeguards to consumer and small-business security
• New York Gov. Hochul signs AI chatbot safety bill into law, protecting kids and communities — Transparency Coalition. Legislation for Transparency in AI Now.
• Gov. Hochul signs bill to shield children from unsafe AI chatbots, baby monitors
• Constraining the Commit Boundary: A DIO/ODI-Informed Authorization Guardrail for Agentic Entity Resolution
• What happens when vendors take the lead on agentic AI
• Anthropic just launched tools to show companies what their AI spending is buying
• Delta Emails Diamond Medallion, Forgets To Delete AI Prompt About Being Warm & Empathetic
• Waters’ new AI tool targets one of the pharmaceutical industry’s most tedious tasks
• CSN lands nearly $3M federal grant to expand AI, tech training
• Critic's Eye: Review-Grounded Graph Reasoning for Long-Form Motif Discovery
• Dageno Launches Free Brand Explorer to Map AI Search Links to Brands, Products, Competitors, and Sources (#97409)