The Washington Post (WP), an American daily, revealed on the 3rd (local time) that 'Google DeepMind', Google's artificial intelligence (AI) division under Alphabet, was found to have included a 'non-disparagement agreement' clause in its employee employment contracts that applies for life even after leaving the company.
Executives and employees are demanding that the clause be deleted, pointing out that it is a toxic clause that prevents them from speaking out publicly about the potential risks of AI even after leaving the company.
WP said it confirmed this fact by reviewing the contract and interviewing two current employees.
The employees interviewed requested anonymity to avoid retaliation from the company.
“The impact of this technology is really big, and the pace of change is very fast,” said one of the incumbents. “The public wants to know and has a right to know. Public statements like this are often the only way for people to learn about the actual and potential harm that is occurring.”
According to one employee, dozens of employees recently participated in discussions about contract language on the company's messaging platform.
The employee reported that some directly criticized the anti-disparagement clause, and others posted emojis containing the word '1984'. '1984' is a very famous novel written by British writer George Orwell (1903-1950), which deals with the horrors of totalitarianism, mass surveillance and surveillance, and ideological censorship.
In recent months, researchers at major AI companies, including Google, OpenAI, and Anthropic, have publicly resigned one after another, citing concerns about the threat that AI could pose to society.
Among these, executives and employees of Anthropic and OpenAI have often publicly expressed concerns about the risks of AI, but WP pointed out that relatively few voices participate in public discussions at Google.
The reason why the non-disparagement agreement became an issue among Google executives and employees was an interview that former Google DeepMind employee Sofia Liguori gave to the British BBC last month.
He has publicly stated his opposition to Google's development of military AI, saying he has a non-disparagement clause in his employment contract but has decided not to be bound by that restriction.
As a result, DeepMind employees in the UK checked their contracts and discovered that they also contained an anti-disparagement clause.
One of the employees said that he did not believe that the 'lifetime no criticism' clause in Google's contract was legally enforceable, and said, "If it is a legally unenforceable clause, the reason for having it is basically to scare people."
There is speculation that Google included this clause to encourage self-censorship among executives and employees.
UK law protects whistleblowers who inform the outside world about company conduct in the public interest.
According to WP, a Google executive responded to an internal discussion regarding the anti-disparagement clause on the 30th of last month and stated that the company would not apply this clause to retirees.
A Google public relations representative said in an email on the 3rd, "We have always encouraged open dialogue about safety and responsibility," and added, "We recently revised the terms and conditions applicable to new employees, including removing the part that required the non-disparagement clause to apply even after the end of the employment relationship."
This representative explained that Google does not apply a non-disparagement clause to former employees, and that there is also a dedicated internal channel for employees to express concerns to top management.
Employees interviewed by WP argued that the company should completely remove the clause not only from new contracts but also from existing employment contracts.
Google initially adopted 'Don't be evil' as its company motto, and was famous for its atmosphere in which executives and employees could openly criticize the company.
The company's management has also promoted welcoming internal differences as Google's proud corporate culture.
In 2018, Google gave up on renewing its contract to provide drone video analysis software to the U.S. Department of Defense due to strong internal opposition, with executives and employees signing a petition opposing it.
However, Google has recently been cracking down on employee protests, and has also removed restrictions related to military AI that were introduced in 2018 in the wake of a petition against it.
In April 2024, Google fired 28 employees the very next day after employees staged protests and sit-ins at its headquarters in New York and Sunnyvale, California, opposing the company's cooperation with the Israeli military.
