Embracing AI in AEC: Striking a Balance Between Benefits and Risks

Are you leveraging artificial intelligence (AI) in your business operations? Have you considered the implications and responsibilities that come with deploying AI technology?

AI technology will affect the way we design, construct, and manage the built environment—whether we want it to or not. From automating repetitive tasks to enhancing decision-making, AI promises significant gains in efficiency, accuracy, and innovation and is already integrated into many of the software tools you use every day. However, like any powerful force, AI also brings with it a host of challenges that must be carefully considered to ensure its responsible and ethical adoption.

As AI continues to integrate into the AEC industry, the importance of having a robust AI policy cannot be overstated. I recently attended an AI conference where I learned about some alarming threats. Did you know that hackers can manipulate AI systems to trigger cyberattacks or tamper with the data used to train AI models? It was quite an eye-opener and highlighted the importance of ensuring employees understand the good, the bad, and the ugly when it comes to using AI tools.

In this Insights piece, I will explore the complex interplay between AI’s benefits and risks, focusing on key areas such as the critical role of human oversight, transparency and accountability, privacy, and fairness. By addressing these issues, firms can unlock AI’s potential while safeguarding their operations and maintaining trust. To stay ahead of the curve, I also provide tips for developing your company’s own policy framework for AI usage. Let’s dive into how such a policy can protect and propel your business forward.

How AI is revolutionizing construction and project management

One of AI’s most significant impacts on construction is its ability to automate repetitive and time-consuming tasks. By employing AI-powered tools, project teams can streamline processes such as material procurement, scheduling, and progress tracking, freeing up valuable human resources to focus on more complex and strategic aspects of project management. This automation reduces the risk of human error and optimizes resource allocation, leading to increased efficiency and cost savings.

AI also plays a pivotal role in enhancing project planning and scheduling. AI algorithms can analyze vast amounts of historical data to identify patterns, predict potential challenges, and optimize project timelines. This enables project teams to make informed decisions, achieve greater project success, reduce delays, and deliver projects on time and within budget.

Also, systems can collect and analyze data on-site by integrating sensors, drones, and Internet of Things (IoT) devices, providing project teams with up-to-date information on various data points. This real-time visibility enables early identification of deviations from the plan, allowing for prompt corrective actions and ensuring projects stay on track.

In addition, AI is revolutionizing the way AEC professionals experience and interact with design development and construction planning. We are already aware of the positive influence of virtual reality (VR) and augmented reality (AR) technologies, but integrating AI has enhanced the immersive experiences that enable users to visualize and interact with project designs in a virtual environment.

Sounds great, right? Read on.

“The More You Know”— AI risks and challenges

AI embodies the quintessential risk-reward paradigm: while it unlocks unparalleled opportunities for innovation and efficiency, it also introduces complex challenges that demand careful navigation. These range from ethical questions and data privacy issues to the potential for biased results and unforeseen consequences. Recognizing and understanding these challenges is the essential first step in implementing AI responsibly. By exploring the nuances of AI risks, we can better equip ourselves to mitigate them and fully leverage this technology’s transformative power. Here are a few to keep top of mind.

The importance of human oversight

While AI holds tremendous potential, it is the skill, experience, and judgment of humans that must steer and supervise its application. Human oversight plays a pivotal role in AI-driven processes and data collection, catching subtle errors that even advanced algorithms might miss. Human involvement not only complements AI’s analytical prowess but also ensures the accuracy and reliability of outcomes by carefully monitoring AI output. Such a balanced approach is instrumental in mitigating legal risks and demonstrating a commitment to responsible and ethical AI implementation.

Transparency and accountability

Without well-defined policies and procedures, AI development and deployment decisions can be inconsistent or unclear, leading to confusion, misuse, and potential harm, including legal and reputational risks for the company.

When integrating AI into your operations, it’s crucial to establish a robust framework for transparency and accountability. This includes implementing policies that ensure clear decision-making processes and define the responsibilities of employees using AI tools in their daily activities.

To promote transparency, empower users by providing clear and concise explanations of how AI systems make decisions. This enables users to understand and question algorithmic outcomes. Additionally, consider regular audits of AI software user agreements to assess performance, identify potential gaps in compliance, and ensure that the terms align with current ethical standards and best practices.  

Privacy, security, and ownership concerns

AI also raises significant privacy, security, and ownership issues.  The current absence of clear AI use regulations and standards creates uncertainty and potential legal liabilities for companies adopting AI technologies. Without a comprehensive framework, companies may struggle to ensure compliance with data protection laws and industry best practices. To complicate matters, privacy laws and regulations may vary by jurisdiction and industry requirements.

Inadequate data security measures further exacerbate privacy and security risks. Without robust safeguards, sensitive information becomes vulnerable to unauthorized access from internal and external sources. This can lead to data breaches, compromising project document confidentiality, intellectual property, and personal information. Ensuring data security throughout its collection, storage, processing, and sharing is critical to maintaining trust and mitigating potential legal consequences.

Lastly, the ownership of AI-generated content can also be a contentious issue. As AI becomes more sophisticated, it can help generate creative works, such as architectural designs, engineering plans, or even construction project proposals. Determining who owns the intellectual property rights to such AI-generated content becomes complex and requires clear policies to address these ownership issues.

Fairness and nondiscrimination in AI algorithms

It is essential to fully understand the potential biases that can arise when interacting with AI algorithms that use diverse datasets. For instance, if an algorithm is trained to recognize horses based on pictures of brown horses and cows based on pictures of black-and-white cows, it may incorrectly identify a brown cow as a horse. ¹ This occurs because the algorithm has associated the color brown with horses and lacks the ability to recognize the distinction unless it is specifically programmed to do so. 

Regular audits are necessary to identify and address any biases that may emerge. To ensure responsible AI usage, ongoing education, and training for professionals are vital. Equipping individuals with knowledge of potential risks and biases enables them to make informed decisions and use AI tools ethically.

Advice on implementing your company’s own AI policy

So, have I convinced you that your company should invest time and resources into developing a comprehensive AI policy? Here are the crucial steps to be taken to implement your company’s own policy effectively.

First, clearly define the purpose and objectives of AI usage within the company. This will serve as a guiding principle for all AI-related activities and ensure alignment with the company’s overall goals and values.

Second, establish clear roles and responsibilities for AI software usage and monitoring. Assign specific individuals or teams to oversee these tasks and ensure accountability throughout the process.

Next, implement robust data governance and privacy practices. This involves establishing protocols for data collection, storage, access, and usage. To safeguard sensitive information and ensure compliance with relevant data protection regulations and industry or contractual requirements. Monitor AI performance within existing software to identify biases, errors, or unintended consequences.

Lastly, AI policies should be regularly reviewed and updated to keep pace with evolving technology, industry trends, and regulatory changes. I encourage you to explore training and webinars discussing AI applications in your industry to stay current on usage, legal implications, and best practices. There are many sources to start your journey of understanding, including ISO global standards² and the U.S. Department of Commerce National Institute of Standards and Technology³ AI standards.

So much of what I have written today may be outdated in just a few months, so stay informed and continue to adapt as your company embraces this powerful tool. 

Eddie Cardozo
Director of Technology

Sources:

1. https://designingrisk.com/2023/01/03/trust-and-bias-in-ai-public-sentiment-and-regulatory-frameworks/
2. https://www.iso.org/obp/ui/en/#iso:std:iso-iec:42001:ed-1:v1:en
3. https://www.nist.gov/artificial-intelligence/ai-standards