The AI Risks Insurance Brokers Need to Advise Clients About

AI Is Expanding the Digital Attack Surface for SMEs and Households

AI is being adopted quickly, but rarely with the same level of governance as traditional systems. 

According to Cisco’s 2025 Cybersecurity Readiness Index, 86% of American business leaders with cyber responsibilities reported at least one AI-related incident over the past 12 months. 

Small businesses are integrating AI into marketing, customer service and operations without fully understanding how data is being handled or where decisions are being automated. Individuals are interacting with AI through financial apps, messaging platforms and everyday online tools. 

Each of these touchpoints expands the digital attack surface, creating new “digital doorways” cyberattackers can exploit.   

Many clients may not know where AI is being used or integrated across their business or household, which makes it difficult to assess exposure. 

Common AI Risks for Small Businesses

Last year, 80% of American small businesses experienced at least one cyberattack yet only 34% of small businesses have a formal incident response plan. 

AI-related losses often show up as familiar incidents, triggered in different ways. 

Intellectual Property and Content Liability 

Businesses are using generative AI to create marketing content, proposals and communications. If that content infringes on existing work, liability sits with the business, not the AI provider. This risk is increasing as organizations rely on AI outputs without reviewing ownership or source material. 

Errors, Hallucinations and Business Liability
 

AI tools can produce incorrect or misleading outputs. When used in customer interactions or decision-making, this can lead to financial loss or liability. These incidents don’t involve system failure, but reliance on flawed output, which is not always clearly addressed in traditional policies. 

Algorithmic Bias and Discrimination
 

AI systems can reflect bias in their training data. When used in hiring, screening or lending decisions, this can create exposure to discrimination claims and regulatory scrutiny, often without businesses fully understanding how decisions are being made. 

Data Privacy and AI Usage
 

Employees are using AI tools to process information and potentially entering sensitive data into public platforms without understanding how it is stored or reused. This creates exposure to privacy breaches, contractual violations and regulatory issues. 

Social Engineering Scams Are Evolving with AI

AI is also changing how attacks are carried out. 

The most significant shift is in how fraudsters target people rather than systems, exploiting trusted relationships. 

Deepfake and Voice Cloning Fraud 

AI can replicate voices using minimal source material.  

This is being used to impersonate executives, request urgent payments and bypass internal controls.  

The request sounds legitimate, which makes it far more difficult to detect. 

 AI-Driven Phishing 

Phishing messages are now more tailored and convincing. 

A recent national fraud survey found that 82% of Americans believe AI is making scams harder to spot. 

AI allows attackers to mimic tone, context and timing, removing many of the signals that once made phishing easier to identify. 

These attacks rely on trust and urgency, not technical sophistication. 

Where Coverage Gaps Are Emerging

AI-related incidents might not trigger traditional coverage. That’s because they don’t always involve a system breach or unauthorized access. 

Instead, losses can result from: 

• voluntary actions based on deception 
• incorrect AI-generated outputs 
• misuse of data through AI tools 

Examples include a payment made after a deepfake call, liability from incorrect chatbot responses or intellectual property claims from AI-generated content. 

These are real losses, but they can fall into grey areas of coverage. Clients often assume they are protected until a claim reveals otherwise. 

Despite growing awareness of deepfake threats, only 29% of American firms have taken steps to protect themselves, with 46% lacking any mitigation plan at all. 

Why Clients Need AI-Specific Cyber Insurance Policies

As AI changes how losses occur, how policy structure keeps pace becomes essential. 

Cyber policies like Cyberboxx® Business are designed as an all-in-one cyber insurance and protection solution, combining comprehensive coverage with always-on preventive services and tools and 24/7 human expert breach response. 

This includes protection for: 

• social engineering and financial fraud 
• data breaches, data exposure and privacy incidents 
• ransomware and cyber extortion, including data encryption and threat-based attacks 
• business interruption 
• incidents involving cloud providers and third-party vendors across the supply chain 

It also includes always-on Cyberboxx Assist® services to help predict and prevent AI-driven and evolving cyber threats, including: 

• attack surface management to identify vulnerabilities 
• dark web monitoring to detect exposed credentials 
• credit monitoring to identify financial exposure 
• access to cybersecurity expertise for guidance and response 

These services help identify risks early and reduce the likelihood of incidents escalating into losses. 

Why BOXX covers Each and Every Loss in a Full Policy Term

Each Cyberboxx® Business policy is also embedded with BOXX’s First Party Each and Every Loss structure, which reinstates policy limits after each separate cyber incident in the full policy term. 

This is important because AI-driven attacks are often not isolated events. A business may experience multiple losses from a single incident or be targeted again during recovery. First Party Each and Every Loss protects clients against multiple cyber incidents for the full policy term. 

If an incident occurs, BOXX Hackbusters® incident response team provides immediate support to contain the threat, investigate the cause and guide recovery. Early intervention can significantly reduce the overall impact. In fact, Hackbusters prevent over 80% of incidents from escalating into insurance claims. 

The Role of Next Gen Tech E&O in AI Risk

For technology companies and AI-enabled businesses, exposure extends beyond cyber. 

This is where next generation Tech Errors & Omissions (E&O) coverage plays an important role. 

Tech E&O by BOXX is integrated with Cyberboxx® Business, combining professional liability coverage with cyber protection. 

It is designed to respond to: 

• AI and large language model errors 
• data misuse and integrity issues 
• technology discrimination claims 
• software and service failures 

As more businesses build or rely on AI tools, this type of coverage becomes increasingly relevant. 

Proactive Steps: Advising Clients on AI Best Practices to Reduce Risk

AI risk can’t be managed through insurance alone. Clients also need clear, practical guidance on how these tools are used day to day. 

Brokers can add immediate value by focusing on a few key areas: 

Set clear boundaries around AI usage
 

Clients should establish guidelines and policies for how AI tools are used, particularly around sensitive data, including restricting the use of public platforms for confidential information. 

Understand third-party AI providers 

Clients need to understand how AI platforms handle data, what rights they retain and where liability sits, especially in customer-facing or decision-making processes. 

Train employees to recognize AI-driven scams 

Less than 25% of US small businesses conduct regular cybersecurity training for their employees, leaving them exposed to AI-generated scams and social engineering attacks. 

Employees should be aware of deepfake voice scams, highly personalized phishing emails and unusual payment or credential requests. 

Strengthen verification controls for financial transactions
Encourage secondary verification for payment requests, banking changes and urgent financial instructions to prevent fraud. 

Review coverage alongside evolving exposure 

Clients should understand what their policies respond to, where gaps may exist and how cyber and professional liability coverage work together. 

Where Brokers Add Value in AI-driven Cyber Risk

AI is changing how losses happen, but many clients are adopting these tools without fully understanding the exposure. 

To strengthen their advisory role, brokers need to understand where AI is creating exposure and where existing coverage may fall short. 

Solutions like Cyberboxx® Business, supported by Cyberboxx Assist® and Hackbusters®, allow brokers to close AI-specific coverage gaps and deliver protection that reflects how losses actually happen today.