AI at Devereux: Approved use, emerging risks – and staying safe

July 28, 2026

Artificial Intelligence (AI) is transforming the way we work, collaborate and solve problems. Unfortunately, cybercriminals are embracing AI as well. Traditional attacks such as phishing emails, fake websites, scams and impersonation attempts are becoming far more convincing because AI can generate realistic content in seconds.

How AI is transforming cyber threats

In the past, suspicious emails often contained spelling mistakes, poor grammar or obvious red flags. Today, AI can create polished messages that closely mimic executives, co-workers, vendors and trusted organizations. AI also can generate realistic videos, voice recordings, text messages and phone calls designed to persuade employees to share information, transfer funds or click malicious links.

Just because something looks, sounds or reads professionally does not mean it is legitimate.

How to spot AI-generated content

Be cautious when you encounter:

• Unexpected requests for sensitive information, payments, gift cards or urgent action.
• Emails that create pressure, fear or a sense of urgency.
• Phone calls or voice messages that sound familiar but request unusual actions.
• Videos or recordings that seem authentic but come from unverified sources.
• Information that cannot be confirmed through official Devereux channels.

When in doubt:

• Verify through a second channel. Call the person directly using a known phone number.
• Check the source. Confirm the sender, website or organization is legitimate.
• Pause before acting. A few extra minutes can prevent a costly mistake.

AI can be wrong

AI-generated content may appear authoritative and confident while still being incomplete, misleading or inaccurate. Always validate information against credible sources and verify facts for accuracy and completeness before using or sharing AI-generated content. Important decisions should never rely solely on AI output.

Approved AI use at Devereux

To protect the data of Devereux, the individuals and families we serve, and our employees, Microsoft Copilot accessed through your Devereux work account is currently the only enterprise-wide approved AI tool. Copilot provides security, privacy and governance controls that align with Devereux standards.

Additional AI solutions for specific business use cases are actively being developed and piloted. More information will be shared as approved tools become available.

Reminder: Per Policy 727 Acceptable Use of Technology, any use of AI tools other than approved enterprise solutions requires prior authorization. Employees should not enter Devereux, employee, individual, financial or other sensitive information into non-approved AI platforms. Use AI as an aid, not a substitute, for professional skills, critical thinking and informed decision-making.

Final reminder

Pause, think and verify before you act. As AI continues to improve, our best defense remains strong cybersecurity habits, healthy skepticism and the use of approved technology solutions. Staying vigilant helps protect our individuals, colleagues