Practitioners Perspectives: GenAI Risks and Opportunities

Generative AI (GenAI) is creating new opportunities, and companies need to keep up to stay ahead. But how can we get our arms around GenAI’s impact on all facets of our organizations? And how do we make the most of GenAI’s potential while avoiding the pitfalls?

We decided to tap into the power of Georgian’s community — The Growth Network — to get a pulse on how tech executives are approaching GenAI at their organizations. The Growth Network is our private community for Georgian customers, with over 2,000 members. The members self-organize working groups to learn from their peers and co-create solutions to shared challenges. The working groups are organized according to function, such as Chief Product Officers, Chief Financial Officers, or around a specific topic, such as employee belonging. 

By bringing together executives from across the Georgian network, our aim for this GenAI initiative was to help executives assess GenAI’s challenges and opportunities. To that end, members of our working groups pulled together to co-create this summary, which includes:

• Opportunities and risks of GenAI specific to each department 
• Chances to collaborate with other teams with the aim of capitalizing on opportunities and mitigating risks
• Areas in the organization where members are aligned, as well as spots where more alignment may be needed

“We want to shape our thinking about this as much as we can at the beginning—so we can help the business embrace this new technology and take well-educated and reasonable risks with it.”

Member of the Legal Officers group

Key Findings

Our community sees both sides of GenAI — it has the potential to be a transformative technology that evokes both promise and trepidation. It holds potential, yet calls for a deliberate approach to mitigate its threats.

1. Leaders report that GenAI tools have the potential to transform every function within our companies.

Here are some of the use cases where our community is using GenAI tools:

1. Policies and guidelines are needed to manage the risks associated with GenAI adoption.
   1. The responsible use of GenAI should be a consideration in addressing risks and seeking to  avoid negative consequences, such as:
      1. Inaccuracies
      2. Copyright infringement
      3. Perpetuation of biases
      4. Elimination of the human touch
      5. Potential erosion of experience and expertise

2. Collaboration and communication across departments is important to effectively implement and leverage GenAI tools.
   1. Legal, Product and Security teams are good resources for leading on policies and guidelines.

3. Senior leadership teams have a responsibility to set the tone and provide reassurance to employees about the role of GenAI in their organizations.
   1. Leaders should give careful consideration to the impact on employee morale, workforce dynamics, human-centric experiences and brand quality when adopting GenAI in different departments.

4. Data and privacy concerns – with vendors and our own products – remain.
   1. Choosing the right partners, navigating terms and conditions, contracts and security concerns, using customer data responsibly, and protecting intellectual property and code are important aspects of implementing GenAI. 

Areas to consider prioritizing

According to a report from McKinsey, “about 75% of the value that GenAI use cases could deliver fall into four areas: customer operations, marketing and sales, software engineering and R&D.”

Potential gains in Engineering productivity:

• 20-45% increase in software engineering productivity 
• better work experience and fulfillment

Potential gains in Customer Support and Service

• 30-50% potential productivity gain of customer operations
• Increase customer satisfaction by providing personalized responses and reducing time to response

Potential Gains in Marketing and Sales

• Increase the productivity of the marketing function with a value 5-15% of total marketing spending
• Increased conversion rate by creating personalized messages 

Poll Results*

Senior leaders from seven of the nine participating groups from the Georgian Growth Network were polled in Q2 of 2023 on their department’s use of GenAI. Here’s the status they indicated:

*This poll includes data from the Georgian Network groups listed above, with the exception of the CMO or CFO groups for which poll data was not available. Data as of June 27, 2023.

What are the top GenAI opportunities you see for your function?

Members of each of the nine responding GN groups were asked to give their views on the ways GenAI could enhance their roles or provide an advantage. Their aggregate answers are below.

Chief Product Officers

• Enhancing data analysis: Chief Product Officers anticipate leveraging GenAI to analyze large volumes of data and try to uncover insights that can inform product strategy and decision-making.
• Boosting ideation and innovation: Members see GenAI’s potential to inform new product ideas, recommend innovative features and suggest novel solutions to customer pain points.
• Streamlining market research: Group members are interested in leveraging GenAI to gather and analyze market data, competitor intelligence and customer feedback.
• Predicting customer behavior: Chief Product Officers discussed using GenAI to predict user behavior, preferences, and needs, enabling personalized product recommendations and improved customer experiences.
• Generating current user feedback: AI-powered sentiment analysis and natural language processing can allow Chief Product Officers to quickly identify trends, issues and opportunities for product improvement.
• Creating personalized product experiences: Leveraging GenAI algorithms, Chief Product Officers can create personalized product experiences tailored to individual users.
• Enabling risk assessment and mitigation: Members foresee tapping GenAI to help identify potential risks and vulnerabilities in product development.

Talent Acquisition

• Automating candidate sourcing: Members foresee using GenAI to find potential candidates by leveraging algorithms to search databases and online platforms — including Workday for internal candidates.
• Screening resumes: Group members are interested in using GenAI to analyze resumes, extract relevant information and match candidate qualifications with job requirements.
• Augmenting candidate engagement: Members mentioned using GenAI to communicate with candidates through chatbots or automated messaging systems, enhancing engagement and providing timely updates.
• Conducting skill and competency assessments: The TA group is interested in exploring how GenAI tools can assess candidates’ skills and competencies through automated tests or simulations.
• Detecting and mitigating bias: Members see GenAI’s potential for flagging biases in job descriptions, promoting fair and inclusive hiring practices.
• Tapping predictive analytics for hiring outcomes: Members are interested in learning how GenAI can help predict candidate success and retention rates

“A pressing thing for us is using GenAI to bridge the gap between Workday and the recruiting system. How can we help people identify roles that they may not have thought about if they are already in the company?” 

Member of the Talent Acquisition group

Chief Financial Officers

• Generating financial report data: CFOs see the potential for using GenAI to generate financial data and produce outputs, using external plug-ins to enhance accuracy.
• Automating cash management: The CFOs discussed GenAI applications that have the potential to revolutionize the finance function — and noted that Trovata, a cash management automation platform, is emerging as an attractive finance application.
• Optimizing market research: The CFO group is interested in exploring how GenAI could enhance their market research capabilities.
• Accelerating rote financial functions: CFOs expect to use GenAI tools to speed standard tasks—especially financial planning and analysis activities.
• Analyzing large data sets: CFOs are most intrigued by using GenAI on large data sets — both standalone and disparate data sets — to conduct predictive analytics, anomaly detection and more advanced data analyses.

️Legal Officers 

• Providing guidance: Legal Officers agree that they will be called on to guide the company on GenAI usage, risk management, policies and contracts.
• Boosting contract efficiency: The Legal Officers group sees the potential to save time and increase efficiency by using GenAI in contract creation, review and redlining. While no one in the group reported using it yet, there is interest about IronClad, which has integrated GenAI.
• Automating tasks: The group sees multiple ways to use GenAI to tackle their rote and repetitive tasks and free them up for more creative and strategic work — from summarizing research and answering routine questions to writing disclaimers and creating terms and conditions.

People Ops/HR Leaders

• Augmenting tasks: Group members agree that leveraging GenAI tools to augment their day-to-day processes could help enable them to focus on employee interactions and more strategic work.
• Supporting recruitment and talent acquisition: People Leaders foresee using GenAI to assist in identifying leading candidates by analyzing resumes and scheduling interviews, in an effort to reduce recruitment time and cost and improve the quality of hires.
• Creating development plans: People Leaders see the potential for using GenAI in learning and development to identify skill gaps and recommend suitable training and development programs.
• Enabling performance management: Leaders mentioned potentially using GenAI to provide feedback and coaching to improve employee performance.
• Testing for inclusive language: With its ability to follow style guidelines, GenAI can help People Leaders streamline the process of reviewing job descriptions, emails and other written communications for inclusive language.

“If we’re using GenAI to automate our role, the bar is so high in terms of performance measurement and error tolerance. But if we’re using it for augmentation there’s… the opportunity for a productivity gain.”

– Member of the People Ops/HR leaders group

Chief Technology Officers/Engineering

• Boosting developer productivity: Most engineering teams are already using GitHub CoPilot for coding suggestions and expect increased productivity in areas like documentation writing, test creation and code language translation.
• Facilitating knowledge management: Some members of the CTO Engineering Group are excited about leveraging Large Language Models (LLMs) to extract and interact with practioners and customers across platforms, creating a centralized knowledge base. Technologies like LangChain and BotPress are being explored for this purpose.
• Enabling multimodal content creation: The Group is interested in utilizing GenAI to generate new content based on existing media, such as images, videos and audio — enabling the creation of personalized content and virtual reality training scenarios. 
• Testing APIs: Group members are curious about leveraging GenAI to produce tests for APIs, in an effort to make testing more efficient, reliable and scalable.

Mergers & Acquisitions

• Staying on top of laws and regulations: Group members expect increased government scrutiny of GenAI and anticipate the need to monitor and comply with new regulations.
• Enhancing GenAI skills and talent: M&A group members foresee the opportunity for companies to bring in new talent with expertise in GenAI through “acqui-hiring” — combined with upskilling existing team members — to enhance the organization’s resources and capabilities.
• Striking the right balance: Some members noted that companies should take an aggressive approach toward experimenting with GenAI, while putting proper guardrails in place.
• Collaborating to innovate: Group members believe that enabling teams to use GenAI effectively—and aggregating results—can lead to valuable contributions and innovations within the organization.
• Patenting algorithms: Group members remarked that innovative algorithms that are core to a company’s technology may be worth protecting through patents—and noted that strategic acquirers may place a high value on patented algorithms.

Chief Marketing Officers

• Enhancing workflows: CMOs are experimenting with multiple marketing workflows that can be enhanced with GenAI—including search engine optimization (SEO), content development, blog writing, code generation for growth marketing, job description writing and business development research.
• Increasing personalization: Group members remarked on GenAI’s ability to create highly personalized messages that sales and business development representatives can use to reach out to customers.
• Rethinking customer discovery: CMOs highlighted the need for marketers to think differently about how customers find and learn about their company, noting the shift from traditional search engines like Google to using chatGPT.

CISOs

• Streamlining security questionnaires/vendor risk assessments: CISOs noted that these are challenging and time-consuming — and are exploring GenAI-based solutions, such as Vendict, which uses GenAI to help automate security questionnaire responses.
• Collaborating with legal: The group expressed a point of overlap with the legal department — and sees a prime opportunity to collaborate on policies.
• Upskilling security engineers: CISOs see GenAI’s potential as a training tool, such as using ChatGPT to propose a one-month/two-hour per day training program on GenAI concepts and technologies.

What GenAI-related risks are you concerned about?

Members of each group were asked to name the threats or pitfalls GenAI could pose. Their aggregate answers are below.

Chief Product Officers

• Ethical considerations: Chief Product Officers are concerned about algorithmic biases, fairness, transparency and the responsible use of data.
• Privacy and data security: Group members note that the collection and analysis of large amounts of user data raises concerns about data privacy, protection and compliance with data regulations.
• Over-reliance on GenAI: Members flagged that GenAI without human judgment may lead to suboptimal decisions, loss of creativity and lack of understanding of customer needs.
• Trust: The group foresees that the opacity of GenAI algorithms and the difficulty in interpreting their decision-making processes may hinder trust.
• Unforeseen biases and unintended consequences: Members reinforced that GenAI models can inherit biases from training data or generate unintended outcomes.
• Impact on workforce and job displacement: The group anticipates that GenAI will necessitate reskilling, upskilling and thoughtful workforce management for product teams.

Talent Acquisition

• Misleading outputs: Members noted that GenAI results can include inaccurate or made-up information, necessitating careful evaluation of results.
• Privacy concerns: The TA group mentioned the need for privacy policies for GenAI tools like ChatGPT to protect confidential data from leaks or unauthorized access.
• Ethical considerations and bias: Group members stressed the need for addressing biases in GenAI algorithms to ensure fair and unbiased candidate selection processes.
• Lack of transparency and interoperability: Members discussed the need for transparency in GenAI models and systems, as well as the ability to seamlessly integrate different GenAI tools and platforms.
• Candidate experience and human touch: TA group members underscored balancing the use of GenAI with maintaining a human-centric approach to try to ensure a positive candidate experience.
• Job displacement: Members recognize the potential impact of GenAI on their roles — and the need to prepare for potential changes to the talent acquisition field.
• Legal considerations: Members discussed the use of GenAI in talent acquisition, including compliance with data protection regulations and intellectual property rights.

Chief Financial Officers

• Accuracy limitations: CFOs noted that while GenAI models provide valuable outputs, their accuracy may not be guaranteed, and caution should be exercised when relying on the generated information.
• Subscription and plugin access: The group discussed that GPT-4 requires a subscription, and access to plugins may be obtained through a waitlist, which could pose limitations in terms of cost and availability.
• Evolving landscape: The CFO group noted that the field of LLMs is rapidly evolving, with new competitors and open-source alternatives emerging. Staying updated and adapting to these changes may be necessary to remain competitive.
• Data sensitivity and privacy: The discussion emphasized the importance of data sensitivity and privacy when using GenAI tools — highlighting the need to implement appropriate measures to safeguard confidential information.
• Limitations in fact-finding and analytics: While GenAI models excel in generating text and filling in information, they may not yet be proficient in fact-finding or advanced analytics.

Legal Officers

• Confidentiality breaches: CLOs are concerned about employees putting confidential information into GenAI systems and recognize that training will be paramount. They also see risks related to violating confidentiality obligations if vendors use confidential information on GenAI platforms — and are considering tweaking standard vendor contracts.
• Copyright ownership: CLOs see the potential risk of copyright infringement claims due to unclear copyright of GenAI output.
• Increasing regulatory focus: As businesses become more dependent on GenAI, CLOs foresee the need to comply with new government-mandated regulations and restrictions.
• Reinforcement and perpetuation of biases: CLOs identified risks related to biases, racism or sexism being perpetuated and reinforced through GenAI results.
• Inherent limitations in predictive language capabilities: CLOs surfaced that employees may not know how GenAI tools work, which could expose them to decision-making based on wrong answers and misinformation.

People Ops/HR Leaders

• Propagating hallucinations and misinformation: People Leaders are concerned that employees may believe that GenAI-generated information is inherently “the truth,” which could lead to misinformed decision-making.
• Eliminating the human touch: Some leaders see a risk in implementing potentially generic GenAI-based interactions that could make employees feel “less important, less valued or less seen.”
• Eroding experience and expertise: People Leaders underscored the years of first-hand experience that makes them highly qualified and effective—and expressed concerns that managers may rely too heavily on GenAI for people-related issues and bypass HR altogether.
• Choosing the right tools: People Leaders stressed the importance of evaluating GenAI-enabled HR tools according to privacy, bias, security and performance measurement criteria.
• Testing for inclusive language: With its ability to follow style guidelines, GenAI can help People Leaders streamline the process of reviewing job descriptions, emails and other written communications for inclusive language.

Chief Technology Officers/ Engineering

• Legal and data security concerns: CTOs are concerned about sending sensitive data to GenAI platforms — particularly because third-party hosts are not providing robust privacy guarantees. They are exploring differential privacy techniques and other ways to add guard rails to these models.

“Obviously, everyone’s really eager to use [GenAI]. But you’re potentially sending very sensitive information—and there are literally no guarantees here. It’s a big gap.” 

Member of the CTO engineering group

Mergers & Acquisitions

• Privacy and security concerns: M&A Group members acknowledged that privacy and security concerns must be addressed while experimenting with GenAI to ensure the protection of sensitive data.
• Negative impact on morale: Group members flagged the need for management teams to address employee concerns and provide reassurance about the role of GenAI — promoting upskilling opportunities to mitigate any potential decrease in morale.
• Ethical considerations: Members underscored the need for organizations to establish clear guidelines and try to ensure responsible and ethical use of GenAI to avoid negative consequences such as privacy breaches or bias.

“Management teams have a responsibility to set the tone and provide reassurance to employees about the role of GenAI in the organization.”

Member of the M&A group

Chief Marketing Officers

• Accuracy of results: CMOs feel that the potential inaccuracy of generated content is a significant risk.
• Trademark and copyright issues: The group sees potential challenges related to AI generated content: Do copyright, patent and trademark infringement apply to AI creations? Is it clear who owns the content that generative AI platforms create?
• Plagiarism concerns: Members use Grammarly for plagiarism checks, but reservations remain about the originality of content created with prompts.
• Impact on other departments: CMOs noted the risk that other departments may feel empowered to create their own content and bypass the marketing department — which could erode quality and impact the brand.
• Legal implications: The group expressed concerns about potential legal issues and the need for policies regarding the use of GenAI.

CISOs

• Choosing the right partner(s): The group listed some exciting small GenAI vendors, but agreed that existing large vendors are a safer bet and may be less time-consuming to manage.
• Navigating contract and security concerns: CISOs and their legal counterparts are concerned that vendor contracts may not offer enough data protection — but are hopeful that ChatGPT for Business may meet enterprise security needs.
• Using customer data: CISOs face uncertainty related to using customer data with ChatGPT and would like to tap guidance from their legal teams.
• Protecting IP and code: The group raised risks related to IP and code development, including the need to disable telemetry.