AI News for the Week of September 18: Updates from Agentic AI Foundation, Cisco, EY & More

Solutions Review Executive Editor Tim King curated this list of notable artificial intelligence news for the week of September 18, 2026. Keeping tabs on all the most relevant artificial intelligence news can be a time-consuming task. As a result, our editorial team aims to provide a summary of the top headlines from the last week in this space. Solutions Review editors will curate vendor product news, mergers and acquisitions, venture capital funding, talent acquisition, and other noteworthy artificial intelligence news items. For early access to all the expert insights published on Solutions Review, join Insight Jam, a community dedicated to enabling the human conversation on AI. AI News for the Week of September 18, 2026 * * Abacus Launches AI Initiative To Help Regulated Enterprises Move From Pilots To Production Abacus has launched a new AI initiative focused on accelerating adoption in regulated industries, combining implementation support, governance, and domain-specific capabilities for organizations with strict security, compliance, and operational requirements. The program is designed to help enterprises operationalize AI while retaining the controls needed for high-stakes use cases. Read on for more→ Abacus Launches AI Initiative to Accelerate Enterprise AI Adoption for Regulated Industries Acquia Finds CMOs And CIOs Lack Alignment On AI Ownership And Governance Acquia’s global survey of 500 CMOs, CIOs, and CTOs found that marketing and IT leaders often disagree over who makes CMS and digital-experience platform decisions, with 57% of CIOs and CTOs describing themselves as the primary decision-maker while marketing respondents report a much more distributed model. The study also found that only 41% of CMOs are confident in their teams’ ability to embed governance into AI initiatives and 61% say content is spread across too many systems to provide a reliable source of truth. Read on for more→ Acquia Research Finds CMOs and CIOs Don’t Agree on Who Leads Their Shared AI Decisions Agentic AI Foundation Launches First Vendor-Neutral Certification For MCP Skills The Agentic AI Foundation has introduced the Model Context Protocol Associate certification, a vendor-neutral credential for developers, platform engineers, and AI-governance professionals working with MCP-enabled systems. The 120-minute proctored exam covers protocol fundamentals, architecture, execution, security and governance, and ecosystem use cases, providing employers with a standardized way to assess knowledge of the fast-growing protocol for connecting agents to external tools and services. Read on for more→ Agentic AI Foundation Launches MCPA Certification to Validate MCP Expertise Airrived Launches Agent Observability For Enterprise Governance, Data Risk, And Cost Control Airrived has introduced Agentic Observability for its Agentic OS, giving enterprises visibility into agent ownership, permissions, human approval requirements, data access, actions, business outcomes, and AI consumption costs. Built around the company’s Context Lake, the capability traces workflows from source data through decision and execution, including the movement of PII, payment-card data, and protected health information, and supports on-premises, private-cloud, and air-gapped deployments. Read on for more→ Airrived Launches Observability for Enterprise AI Agents Cisco Extends Splunk AI To Private And Air-Gapped Environments With Agent Controls And Token Economics Cisco has added an AI POD for Splunk to its Secure AI Factory with NVIDIA, enabling Splunk Enterprise customers to run AI Assistant, forthcoming Agent Launchpad, and selected open or proprietary models inside on-premises, private-cloud, and air-gapped environments. Cisco also expanded Splunk Agent Observability across Splunk Observability Cloud and Cisco Cloud Control with Tokenomics for real-time AI-cost attribution and forecasting, while new agentic SOC capabilities, Exposure Analytics, and a multi-year AWS co-development agreement target faster, governed security operations. Read on for more→ Cisco Delivers Trusted AI at Scale Through New Splunk Advancements Collibra Finds AI Teams Are Paying A “Hallucination Tax” To Validate Agent Outputs Collibra’s Harris Poll survey of 306 U.S. data, privacy, and AI decision-makers found that 72% attribute underperforming AI initiatives to weak or unaligned data foundations, while 76% have encountered critical roadblocks moving agent programs from pilot to production. Eighty-seven percent regularly re-verify the accuracy and currency of agent context, and 51% spend significant staff time reviewing and correcting agent outputs before release, underscoring how incomplete governance and context turn promised automation into manual oversight. Read on for more→ New Survey From Collibra by The Harris Poll Finds 72% of Tech Decision-Makers Feel AI Initiatives Today Are Falling Short DataCebo Releases Generative Relational Models For Privacy-Safe Enterprise AI Development DataCebo has launched SDV 2.0, enabling organizations to create generative relational models that learn the structures, relationships, rules, and statistical patterns across enterprise databases inside customer-controlled infrastructure. The models can generate synthetic data, simulate rare scenarios, and provide context for AI-agent training and evaluation without exposing production data; SDV 2.0 automates schema and relationship discovery, business-constraint detection, model tuning, and task-specific dataset creation across sources including Oracle, SQL Server, BigQuery, Spanner, and AlloyDB. Read on for more→ DataCebo Releases SDV 2.0 for Building Generative Relational Models of Enterprise Data EY Finds Autonomous AI Deployments Are Moving Faster Than Enterprise Governance EY research finds that organizations are accelerating the rollout of autonomous AI capabilities while governance, accountability, and oversight frameworks struggle to keep pace. The resulting gap increases the importance of defining decision rights, monitoring agent behavior, managing access to sensitive data and systems, and establishing escalation paths before autonomous systems take on broader business responsibilities. Read on for more→ EY Survey: Autonomous AI Implementation Outpaces Oversight, Yielding an AI Governance Gap Expanse Raises $5.3 Million To Predict AI Compute Needs Before Workloads Run Expanse has raised $5.3 million in seed funding led by Crane Venture Partners to expand software that predicts the optimal GPU, CPU, memory, and runtime requirements for AI workloads before infrastructure is allocated. Deployed inside customer cloud or on-premises environments, the platform aims to improve utilization, prevent failed jobs, reduce over-provisioning, and keep code and telemetry within customer boundaries; Expanse says one production deployment identified nearly $8 million in idle compute capacity in a month. Read on for more→ Expanse Raises $5.3 Million to Eliminate Guesswork From AI Computing Factory Raises $200 Million To Build Governed, Self-Improving Software Development Systems Factory has raised $200 million at a $5 billion valuation, more than tripling its $1.5 billion valuation from April and bringing total funding above $400 million. The platform enables enterprises to operate autonomous “software factories” that build and improve software under customer control, with support for managed cloud, on-premises, and air-gapped deployment models; Factory says customers include Nvidia, Blackstone, RBC, Palo Alto Networks, and Adobe. Read on for more→ Factory Raises $200 Million at a $5 Billion Valuation to Scale Self-Improving Software Development in the Enterprise HPE Doubles Alletra Storage MP X10000 Scale For Expanding AI Data Estates HPE has released the fourth software update for Alletra Storage MP X10000, expanding the object-storage platform’s scale and capabilities for data-intensive AI and analytics environments. The update is designed to help