More than 50 state agencies in Maryland already use AI — and transparency is in check

More than 50 agencies, councils and offices of the Maryland government in the United States already use artificial intelligence in their operations, according to a survey released by the GovTech portal. The accelerated adoption, ranging from citizen care chatbots to data analysis tools for public policies, occurs without a unified governance structure, sparking warnings about transparency, accountability and the risk of an institutionalized Shadow AI: the ghost that surrounds your company while employees use personal AI.

Why does Maryland adopt mass AI?

The state of Maryland is not an isolated case, but the pace of adoption surprised experts. Unlike pilot initiatives in other states, Maryland recorded an organic growth: the agencies themselves sought AI tools to solve specific bottlenecks, such as document sorting, forecasting lawsuits and optimization of public transport routes. The absence of a central guideline, however, caused each distribution to adopt different solutions, often without prior evaluation of ethical or safety risks.

What are the differences between AI tools in Maryland?

Applications vary enormously. While the transit agency uses AI to predict congestion, the health department employs machine learning models to identify outbreaks of disease. There are also chatbot systems to answer citizens' questions about benefits and documents. The problem, according to the report, is that most of these implementations did not go through external audit or were not even formally communicated to the population. This creates a scenario similar to what happens in companies, where the computational cost of AI already exceeds the salary of human employees, as Nvidia’s PV admitted — an issue that gains delicate contours when public money is at stake.

How does the lack of transparency in AI threaten public trust?

Without clear rules, citizens do not know if their information is being processed by algorithms, nor do they have the means to appeal to automated decisions. The lack of supervision also opens a gap for biases in models, something particularly serious in areas such as social assistance and criminal justice. In Maryland, the situation is aggravated by a legal vacuum: there is no state law that forces agencies to report the use of AI, which makes the survey of GovTech a voluntary and therefore probably incomplete portrait.

What do experts in AI governance say about transparency?

For public policy scholars, the way out is not to prohibit AI, but to build a governance framework that includes algorithmic transparency, independent audit and citizen participation. Maryland’s experience serves as a warning to other governments — including the Brazilian one — that advance in digitization without the same care. In parallel, initiatives such as the partnership between TCS and Anthropic to take Claude to regulated industries show that it is possible to reconcile innovation with compliance. The challenge is to do this on a state scale, with a tight budget and without losing sight of the public interest.

Frequently Asked Questions about AI in Maryland

What does it mean for a state agency to use AI?

It means that public agencies employ systems based on artificial intelligence to automate tasks, analyze data or make decisions, such as chatbots for service or algorithms to prioritize inspections.

How many Maryland agencies are using AI today?

More than 50 state agencies, councils and departments reported the active use of AI tools, according to a survey by GovTech in April 2025.

What are the main risks of this uncoordinated adoption?

Lack of transparency on automated decisions, potential algorithmic bias, absence of independent audit and use of personal data without adequate consent.

Source: GovTech - More Than 50 Maryland State Agencies Now Using AI