Anthropic co-founder Chris Olah spoke at the Vatican presentation of Pope Leo XIV’s AI encyclical, Magnifica Humanitas, and made a rare point for an AI lab leader: frontier AI companies need serious outside critics because commercial, geopolitical, and competitive incentives can conflict with the public good. For business leaders, the takeaway is clear. AI governance cannot live only inside IT, product, or innovation teams. It needs cross-functional oversight, human impact review, data governance, adoption planning, and clear rules for how AI is allowed to act inside business systems like Salesforce, HubSpot, service workflows, and customer data platforms.
Vantage Point helps organizations evaluate AI through that operating lens: what the workflow is, what data it uses, what risk it creates, who is accountable, and how it should be implemented safely across CRM and revenue operations.
On May 25, 2026, Pope Leo XIV released Magnifica Humanitas: On Safeguarding the Human Person in the Time of Artificial Intelligence. Anthropic co-founder Chris Olah was invited to speak at the Vatican presentation of the encyclical as part of Anthropic’s broader effort to widen the conversation about frontier AI.
Olah’s remarks were notable because he did not present AI safety as something AI companies can solve alone. He said every frontier AI lab, including Anthropic, operates within incentives that may conflict with doing the right thing: commercial viability, research competition, geopolitical pressure, pride, and ambition.
His central point was simple and important: society needs people outside those incentives who are paying attention, willing to say hard things, and able to act as thoughtful critics. That includes religious communities, civil society, scholars, governments, and other groups that can evaluate AI through a human, moral, and social lens.
This matters for companies because enterprise AI adoption is moving faster than most governance models. AI is now being considered for sales prospecting, service triage, content creation, knowledge management, workflow automation, data analysis, customer communications, and agentic task execution. The question is no longer whether AI can do useful work. The question is whether the organization has the structure to decide what AI should do, where humans must remain accountable, and what risks should be controlled before rollout.
AI governance matters in 2026 because AI is moving from experimentation into operational systems. A chatbot in a test environment is one risk profile. An AI agent connected to CRM records, support cases, meeting notes, marketing workflows, or customer communications is another.
Pope Leo’s encyclical emphasizes human dignity, the common good, transparency, responsibility, concentration of digital power, work, freedom, and the impact of technology on vulnerable people. Olah’s remarks add a practical warning from inside the AI industry: the organizations building frontier models face pressures that outsiders should not ignore.
For business leaders, this creates a useful bridge between ethics and implementation. Responsible AI is not only a policy statement. It is a set of business decisions about data, workflows, access, accountability, adoption, and oversight.
| AI governance question | Business operating question | Practical control |
|---|---|---|
| Who benefits from this AI use case? | Does this workflow improve customer, employee, and business outcomes? | Use-case approval criteria tied to measurable process goals |
| Who could be harmed? | Could the AI create bias, exclusion, privacy risk, or bad customer decisions? | Risk review before launch and recurring post-launch monitoring |
| What data does the AI use? | Is the CRM, support, marketing, or integration data accurate and permissioned? | Data quality checks, field governance, access controls, retention rules |
| Who is accountable? | Who owns the decision if the AI takes or recommends an action? | Human-in-the-loop approval and named business owners |
| How will people adopt it? | Will employees understand when to trust, question, or override AI? | Training, enablement, feedback loops, and change management |
| How will it be audited? | Can leaders see what happened and why? | Logging, review queues, escalation paths, and governance dashboards |
The most useful lesson is that AI oversight should include people who are not rewarded only for speed, deployment volume, or short-term productivity gains.
That does not mean slowing every AI project to a crawl. It means creating enough friction in the right places. Before AI is connected to customer data, workflow automation, or decision support, leaders should know what problem it solves, what data it relies on, what actions it can take, and how humans can intervene.
Olah’s comment about the need for “earnest, thoughtful critics” also applies inside companies. The best AI programs include operators, compliance leaders, frontline users, security teams, data owners, customer-facing leaders, and executives. Each group sees different risks.
A sales leader may focus on pipeline efficiency. A service leader may focus on response time. A compliance leader may focus on privacy and auditability. A RevOps leader may focus on data quality and process consistency. A frontline user may see where an AI recommendation sounds right but fails in the real workflow.
Good governance turns those perspectives into a better implementation plan.
AI governance inside Salesforce and HubSpot should start with workflow design, data readiness, and permission boundaries. These platforms hold customer records, lifecycle history, sales activity, service conversations, marketing consent, and operational context. That makes them powerful places to use AI, but also sensitive places to automate without guardrails.
For Salesforce, governance often needs to address role hierarchy, field-level security, sharing rules, data model complexity, integrations, automation overlap, and how AI-generated recommendations appear inside user workflows. Teams evaluating Agentforce, Einstein, Data Cloud, or related AI capabilities should start with a defined use case and a clear data map. Vantage Point’s Salesforce implementation and advisory services help teams connect platform capability to operating model, data quality, and adoption planning.
For HubSpot, governance often needs to address content operations, marketing permissions, lifecycle stages, sales handoffs, service context, and how AI-generated content or recommendations affect customer experience. Teams adopting HubSpot AI capabilities should review source data, content approval workflows, brand standards, and customer communication rules. Vantage Point’s HubSpot consulting services help teams optimize HubSpot across Marketing, Sales, Service, Content, Commerce, and Operations Hubs.
For organizations using both Salesforce and HubSpot, governance must also account for sync logic, field mapping, lifecycle definitions, duplicate prevention, and ownership of shared records. The AI layer can only be as reliable as the data and process underneath it. Vantage Point’s HubSpot and Salesforce integration guidance helps teams design cleaner data flow and clearer accountability between platforms.
Businesses do not need to wait for perfect regulation or perfect vendor guidance before creating practical AI controls. Start with a short checklist that connects ethics to operations.
The practical goal is not to block AI. The goal is to make AI useful, reviewable, and aligned with how the business actually works.
AI governance can fail when companies treat it as a legal policy, an IT security project, or a vendor-selection checklist instead of a business operating model.
If governance is too narrow, teams may approve tools without understanding downstream workflow effects. AI may summarize a support case without knowing the latest customer history. It may generate outreach based on incomplete lifecycle data. It may recommend next steps from outdated CRM fields. It may automate a task that should require human judgment.
The risk is not only that AI makes mistakes. The larger risk is that AI makes existing operational weaknesses faster and harder to see.
That is why data governance, integration architecture, and process ownership matter. Vantage Point’s system integration and data migration services help organizations clean up the foundation before adding more automation. Its compliance and security solutions help teams evaluate access, privacy, and governance requirements before sensitive workflows scale.
Businesses should move from broad AI enthusiasm to a specific AI governance roadmap. The roadmap should identify which use cases are ready now, which require data cleanup, which need compliance review, and which should wait until workflows are better defined.
A practical next step is to choose one high-value workflow and evaluate it end to end:
This approach keeps AI grounded. It also helps leaders avoid two common mistakes: deploying AI broadly before the process is ready, or delaying every AI initiative because governance feels too abstract.
The right middle path is structured experimentation: limited scope, clear controls, measurable value, and fast learning.
Vantage Point helps organizations evaluate, implement, and optimize Salesforce and HubSpot based on their operating model, data needs, adoption goals, and growth strategy. For AI initiatives, that means helping teams connect vendor capabilities to real workflows, clean data, responsible automation, and practical governance.
If your team is evaluating how AI applies to Salesforce, HubSpot, integrations, customer operations, or CRM governance, Vantage Point can help assess the right next step and build a practical implementation plan. Relevant services include AI-driven personalization and analytics, CRM and marketing automation strategy, advisory and change management, and managed services and ongoing support.
Chris Olah said frontier AI labs need informed outside critics because AI companies operate under commercial, geopolitical, research, pride, and ambition pressures. He praised Magnifica Humanitas as part of the broader discernment needed to guide AI toward human benefit.
Magnifica Humanitas is Pope Leo XIV’s 2026 encyclical on safeguarding the human person in the time of artificial intelligence. The document addresses AI through themes including human dignity, work, truth, freedom, power, responsibility, transparency, and the common good.
The encyclical matters to businesses because it frames AI as a human and organizational responsibility, not just a technical opportunity. Companies deploying AI in CRM, service, marketing, or operations need governance that protects people, data, trust, and accountability.
Companies should start by selecting one specific AI use case and mapping the workflow, data, permissions, risks, human review points, and audit requirements. This makes governance practical and keeps AI adoption tied to real business outcomes.
CRM data is central to AI governance because AI recommendations often depend on customer records, lifecycle stages, activity history, consent data, support cases, and sales context. If that data is incomplete, duplicated, outdated, or poorly permissioned, AI can produce unreliable or risky outputs.
AI governance should not be owned by IT alone. IT, security, compliance, RevOps, business leaders, data owners, and frontline users should all participate because AI affects systems, people, customer experience, and business decisions.
Vantage Point helps teams translate AI strategy into CRM, data, integration, security, and adoption decisions. That includes evaluating Salesforce and HubSpot use cases, improving data readiness, designing governance controls, and supporting rollout through change management.