Understand this story
On October 30, 2026, Oracle announced the Deep Trust AI Architecture at Oracle AI World, pushing security and business rules down into Oracle AI Database so that the rules do not depend on whether a specific application or AI agent implements them on its own. According to heise online, the architecture includes Deep Data Security, which verifies user identity using cryptographically signed tokens issued by the enterprise directory; SQL Assertions, which block operations exceeding limits according to business conditions; Deep Trust Data APIs, which let applications and agents perform controlled reads and writes on logical business objects such as orders; and Deep NL2SQL, which improves natural language to SQL conversion. Oracle did not provide a unified launch date or overall pricing, and the availability of each component must be checked separately.
AI-generated from public reports
Latest development · Report
Oracle uses real application testing to ensure the safety of AI-driven security patchesDevelopments
1 confirmed developments
- Source:heise online
Reports and discussion timeline
Browse reports and discussion by time. Their evidence levels remain distinct.
Nov 5
SiliconANGLE
Oracle uses real application testing to ensure the safety of AI-driven security patches
Oracle Executive Vice President Juan Loaiza said at Oracle AI World that the company obtained early access to the Mythos model in March of this year and used it to scan its own code, discovering a large number of serious security vulnerabilities. The fixes were subsequently included in quarterly patches, and one quarterly security update this year fixed more than 1,400 vulnerabilities. Oracle uses Real Application Testing to capture real workloads to validate patches, and through Deep Data Security, delivered starting in April, it passes user identities into the database as encrypted tokens, enforcing security and business rules at the data layer.
Oct 30
heise online
heise Weekly: Mistral Open-Sources ML4, Cohere Secures Investment from Schwarz Group, Google Restricts Gemini…
On Thursday, Mistral released the weights of its largest model, Mistral Large 4 (ML4). The model has a total of one trillion parameters, activates only 49 billion per token, and can run on four GPUs. Mistral says it outperforms Kimi K3 in coding and cybersecurity. Cohere completed its largest funding round to date, led by Schwarz Group, which owns Lidl and Kaufland, with participation from the Canadian government and Nvidia. Google launched Gemini Carbon four weeks after Gemini Argon, but at the White House's request it is open only to U.S. intelligence agencies and select partners. In addition, an Anthropic experiment involving 50 statisticians showed that in 86% of cases experts adopted the option given by an AI agent, and about half treated the final recommendation as their own judgment.
heise online
Oracle launches Deep Trust AI architecture to constrain autonomous AI agents at the database layer
At Oracle AI World, Oracle announced the Deep Trust AI Architecture, which pushes security and business rules down into Oracle AI Database so that the rules do not depend on whether a specific application or AI agent implements them on its own. Deep Data Security verifies user identity using cryptographically signed tokens issued by the enterprise directory, while SQL Assertions block operations that exceed limits according to business conditions; Deep Trust Data APIs let applications and agents perform controlled reads and writes on logical business objects such as orders, and Deep NL2SQL improves natural language to SQL conversion. Oracle did not provide a unified launch date or overall pricing, and the availability of each component must be checked separately.