Teradata rolled out a new suite of security and privacy tools on October 20, 2026, aimed at protecting sensitive data used by AI agents. The updates include poly-anonymization, format-preserving encryption, post-quantum security, and immutable backup protection. Through a partnership with Anonomatic, customers gain access to the Anonomatic PII Vault for privacy protection. These capabilities are available on Teradata Cloud across AWS, Azure, and Google Cloud. Some experts note that competitors like IBM and Oracle already offer similar features.
Sumeet Arora, Teradata's Chief Product Officer, stressed that privacy, security, and resiliency must work together as AI environments grow more complex. He pointed out that AI moves sensitive data across users, applications, and agents simultaneously, making unified protection essential. The new database-level high availability features also aim to strengthen enterprise resilience.
Elsewhere, Creatio's second annual State of Enterprise AI Adoption Report, also released on October 20, 2026, surveyed over 600 IT and business decision-makers globally. The findings show that 70% of organizations have moved beyond AI experimentation, a notable shift from 2025 when 86% of leaders expected AI agents to become important within two to three years. The report explores how companies are scaling AI and what separates advanced adopters from the rest.
ServiceNow published a separate study warning that ungoverned agentic AI is spreading inside enterprises. Surveying over 3,700 security leaders, it found that 90% of organizations experienced an agentic AI security incident in the past year. Only 13% maintain a defensible security posture across five key domains. Alarmingly, 29% of organizations have unknown, unregistered AI agents running in production. ServiceNow launched its Autonomous Security offering to help organizations regain control over AI agent deployments.
On the threat side, a hacker used AI tools to breach more than 100 global companies, spending an average of $50 per intrusion and about $5,000 in total. Cybersecurity experts say AI-powered attacks are becoming more common and harder to detect. Interpol also warned that criminals are using AI to fake faces, voices, and behavior for fraud, impersonating executives or family members to pressure victims into sending money quickly. The agency recommends verifying callers, using code phrases, enabling two-factor authentication, and training employees to recognize social engineering attacks.
In other developments, Richmond, Virginia plans to deploy an Amazon AI system to handle up to 60% of its non-emergency 911 calls, more than triple the current share managed without human dispatchers. Womble Bond Dickinson launched its Womble Artificial Intelligence Lab on the same date, focusing on proprietary AI tools for regulatory monitoring, legal analysis, and document management. The firm piloted AI tools in 2025 with positive results. Meanwhile, Goldman Sachs traders were caught off guard by a sudden market shift involving AI-related stocks, showing that even reliable models can struggle during fast-moving conditions.
Key Takeaways
- Teradata launched new AI security features including poly-anonymization, format-preserving encryption, and post-quantum security, available on Teradata Cloud across AWS, Azure, and Google Cloud.
- Creatio's survey of 600 decision-makers found 70% of organizations have moved beyond AI experimentation, up from expectations in 2025.
- ServiceNow's study of 3,700 security leaders found 90% of organizations had an agentic AI security incident in the past year, with only 13% having a defensible security posture.
- 29% of organizations run unknown, unregistered AI agents in production, according to ServiceNow research.
- A hacker breached over 100 companies using AI tools at an average cost of $50 per intrusion, totaling about $5,000.
- Interpol warns that criminals are using AI for deepfake fraud, impersonating executives and family members to pressure victims into sending money.
- Richmond, Virginia plans to use an Amazon AI system to handle up to 60% of its non-emergency 911 calls.
- Womble Bond Dickinson launched its Womble Artificial Intelligence Lab on October 20, 2026, to build proprietary AI tools for legal and regulatory work.
- Goldman Sachs traders struggled when AI-related stocks shifted suddenly, with the firm's AI models unable to keep up with rapid market changes.
- Teradata partnered with Anonomatic to deliver privacy protection through the Anonomatic PII Vault, with Sumeet Arora emphasizing the need for unified privacy, security, and resiliency.
Teradata launches new AI security tools to protect sensitive data
Teradata introduced new security features to protect data used by AI agents. The updates include poly-anonymization, format-preserving encryption, post-quantum security, and immutable backup protection. These tools aim to reduce identity exposure while keeping data useful for analytics. Privacy capabilities from Anonomatic are now available to Teradata customers. Some experts say similar features already exist from competitors like IBM and Oracle.
Teradata unveils privacy and resiliency tools for enterprise AI
Teradata announced new privacy, security, and resiliency capabilities on October 20, 2026. Key features include Poly-Anonymization, format-preserving encryption, post-quantum security, immutable backup protection, and database-level high availability. The company partnered with Anonomatic to deliver privacy protection through the Anonomatic PII Vault. Sumeet Arora, Teradata's Chief Product Officer, said these tools address how AI moves sensitive data across users, applications, and agents simultaneously.
Teradata adds security features to meet enterprise AI challenges
Teradata released new data protection capabilities designed for enterprise AI environments. The update includes poly-anonymization, format-preserving encryption, post-quantum readiness using Anonomatic PII Vault and TLS 1.3, immutable backup protection, and database-level high availability. Privacy tools powered by Anonomatic are available now on Teradata Cloud across AWS, Azure, and Google Cloud. Sumeet Arora emphasized that privacy, security, and resiliency must work together as AI environments grow more complex.
Creatio reports enterprise AI adoption is accelerating
Creatio released its second annual State of Enterprise AI Adoption Report on October 20, 2026. The survey of over 600 global IT and business decision-makers found that 70% of organizations have moved beyond AI experimentation. This is a shift from 2025 when 86% of leaders expected AI agents to become important within two to three years. The report examines how companies are scaling AI and what separates advanced adopters from others.
ServiceNow warns enterprises lack control over agentic AI
ServiceNow published research warning that ungoverned agentic AI is growing inside enterprises. The study surveyed over 3,700 global security leaders and found 90% of organizations had an agentic AI security incident in the past year. Only 13% have a defensible security posture across five key domains. Many organizations run hundreds or thousands of AI agents, with 29% having unknown unregistered agents in production. ServiceNow launched its Autonomous Security offering to help organizations gain control over AI agent deployments.
AI Training Sessions Rescheduled for December 2
AI training sessions have been moved to December 2. The original schedule was changed, and the new date is now set for that day.
Goldman Traders Struggle as AI Market Shifts Suddenly
Goldman Sachs top traders were caught off guard by a sudden market shift involving AI-related stocks. The firm's AI models could not keep up with the rapid changes. Traders were forced to make quick decisions as equities, credit, and derivatives all moved unpredictably. The confusion showed how even reliable systems can struggle during sudden market shifts.
AI Hacker Breaches Over 100 Companies for $50 Each
A hacker used AI tools to break into more than 100 global companies, spending an average of $50 per intrusion. The total cost of all breaches was about $5,000. The report did not name the affected companies. Cybersecurity experts warn that AI-powered attacks are becoming more common and harder to stop. Companies are urged to improve security and use AI-based defense tools.
Interpol Warns About AI-Based Fraud and Deepfake Scams
Interpol warned that criminals are using AI to fake faces, voices, and behavior for fraud. Scammers impersonate executives, friends, or family to pressure victims into sending money quickly. Interpol advises people to verify callers, use code phrases with loved ones, enable two-factor authentication, and report suspected fraud. Companies should train employees to recognize phishing and social engineering attacks.
Womble Bond Dickinson Launches AI Development Lab
Womble Bond Dickinson announced the launch of its Womble Artificial Intelligence Lab on October 20, 2026. The lab focuses on building proprietary AI tools for regulatory monitoring, legal analysis, document management, and contract management. The firm had tested a pilot program for AI tools in 2025 with positive results. The lab is embedded within the firm's IT team and includes engineers who develop technology for the organization.
AI Recipe Sites Surge Along With Ad Tech MFA Concerns
DecryptAds found over 2,100 AI-generated recipe sites that run programmatic ads and use fake author profiles. The rise of these AI slop sites raises questions about whether ad tech is ready for new multi-factor authentication challenges. The sites monetize automated content through standard advertising networks. This trend highlights growing concerns about AI-generated content in the ad ecosystem.
Richmond Considers AI for Non-Emergency 911 Calls
Richmond, Virginia plans to use an Amazon AI system to handle up to 60% of its non-emergency 911 calls. This would be more than triple the current share managed without human dispatchers. The system would process routine calls that do not involve immediate danger or emergency response. The city is exploring AI to manage its growing call volume.
Key AI Terms You Need To Know
As AI becomes more common, understanding key terms is important. Abliteration removes safety guardrails from an AI model so it will not refuse requests. An AI agent uses tools like web browsers and apps to perform tasks such as sending emails or booking reservations. Alignment means training an AI model to behave in line with human goals and values, as defined by OpenAI chief scientist Jakub Pachocki.
Sources
- Teradata's new security measures aim to harness AI agents
- Teradata Introduces New Privacy, Security, and Resiliency Capabilities as AI Forces Change in Enterprise Data Protection
- Teradata Introduces New Privacy, Security, and Resiliency Capabilities as AI Forces Change in Enterprise Data Protection
- Creatio Releases The State of Enterprise AI Adoption Report: An In-Depth Look at Where Enterprise AI Stands
- ServiceNow Research Study Warns Security Leaders: Ungoverned Agentic AI Exposure Is Compounding Inside Enterprises
- RESCHEDULED: AI Training Sessions moved to Dec. 2
- Drowning in the Market Riptide: Goldman’s Top Traders Explain the AI-Sea of Burning Shadows
- AI Hacker Breaches 100+ Global Companies at $50
- Video: Interpol Issues Warning About Fraud Involving Artificial Intelligence
- Womble Establishes Lab for Developing Proprietary AI Tools
- AI Slop Sites Are On The Rise. Is Ad Tech Ready For Its New MFA Challenge?
- Should AI handle Richmond’s non-emergency 911 calls?
- As AI dominates conversations, here are key terms to know
Comments
Please log in to post a comment.