Nearly half of legal professionals spend three hours or more reviewing a single contract, according to a recent survey, leading many in-house counsel and firms to consider whether artificial intelligence contract tools might improve their redlining work.

One-third of the 150 respondents in the survey said it takes 3-5 hours to review a contract, while about 10% said they don’t know how much of their time one contract typically requires, according to LegalOn Technologies, which sells contract-review software.

About 60% of legal teams don’t operate with playbooks, the guidelines that outline contract-review procedures and that help ensure agreements meet company standards, the survey found. These rules also provide guardrails for legal team’ contract reviews.

Most legal departments don’t have the staff time to create extensive playbooks, LegalOn said. Only a quarter said they do have a comprehensive contract playbook; many use a general clause library. 

“Without standardized playbooks, legal teams often have to reinvent the wheel for each review,” according to the report, “The State of Contracting in the Age of AI.” About 76% of the respondents who participated in the survey work as GCs or in-house counsel, about 10% were in legal operations and about 5% were outside counsel.

AI tools may help reduce the staffing and redlining challenges but “AI adoption is still very much in its infancy,” LegalOn said in the white paper. More than 70% of the professionals said their teams are “actively” considering or open to AI usage, while only 8% said AI was being used currently.

One reason for this sluggish adoption is the widespread limitations still endemic to AI use in legal work: Errors are rampant with AI tools built upon large language models. 

A widely cited Stanford University study released in January found that these models produced hallucinations between 69% and 88% of the time when queried about a legal matter. For 200 legal queries in the Stanford study, AI software produced inaccurate information as much as one-third of the time. 

“While AI can speed up time-intensive legal tasks, there is a ‘jagged frontier’ that separates tasks at which AI is more successful from those at which it is less,” the report said.

And while rich, detailed legal research is beyond the ken of most AI software, contract review appears to lie within its capabilities, LegalOn said, citing several factors that differentiate redlining from deeper legal work.

• Defined standards: AI tools can be focused on specific contract types and clauses
• Easier verification: AI contract review analyzes individual documents, making it easier for lawyers to inspect and validate the work
• Focused objectives: Contract review is about finding standard clauses and deviations, two tasks that AI does well with its pattern-recognition skill
• First-pass review: AI can analyze contracts quickly and identify key issues, potential risks, and deviations from standard language
• Context/reasoning: On suggested changes or issues the AI flags, the software “can provide context and reasoning, empowering attorneys to make informed decisions quickly,” LegalOn said.