Energy Cabinet Secretary Opiyo Wandayi has challenged claims linking the Ministry of Energy and Petroleum to an AI-generated social media campaign that sought to portray widespread public support for the government’s decision to withdraw a proposed electricity tariff increase.
Wandayi said there was no evidence showing that he or the ministry approved, directed or took part in the online campaign, which allegedly relied on Artificial Intelligence to produce large numbers of posts praising the government’s position on electricity prices.
The CS was responding to findings contained in a report on coordinated inauthentic behaviour in Kenya. The report described an operation in which an unidentified actor allegedly used Claude, an artificial intelligence chatbot, to create batches of 50 posts that were made to look like genuine comments from ordinary Kenyans.
A significant part of the content was said to have praised Wandayi after the proposed Kenya Power tariff increase was withdrawn. The campaign used hashtags #PowerReliefKE and #PoweringTheNewKenya as part of efforts to spread the messages online.
Despite the pro-government nature of the posts, investigators did not establish who was behind the activity. The report also found no evidence that the government, the Energy Ministry or Wandayi had been involved.
“My attention has been drawn to some unverified report alleging the creation of some activity on social media that is purported to have been in praise of the actions of the Ministry of Energy and Petroleum regarding electricity tariffs,” Wandayi said.
He argued that the existence of anonymous accounts coordinating messages on social media could not automatically be taken as proof that government officials or institutions were responsible for them.
“There are probably millions of anonymous and coordinated accounts on social media pushing different narratives every day. What matters is evidence,” he said.
Wandayi further pointed to the report itself, saying its findings did not establish a direct connection between the alleged operation and the ministry.
“The report itself does not establish that either the Cabinet Secretary or the Ministry authorised, directed, or otherwise participated in the alleged activity.”
The Energy CS said the ministry would continue to explain its programmes and decisions through official communication channels instead of responding to claims based on anonymous online activity.
“We remain focused on communicating the Ministry’s work through credible channels and will not be distracted by speculation around anonymous online activities,” he said.
The report, identified as GTG-54004: Disrupting a domestic coordinated inauthentic behaviour campaign in Kenya, said the operation was designed to make political messaging appear to be coming naturally from members of the public.
The activity went beyond the electricity tariff debate, with the same campaign also pushing claims that the United Opposition coalition was splitting ahead of the 2027 General Election.
Former deputy president Rigathi Gachagua and retired president Uhuru Kenyatta were among the political figures targeted by the campaign.
Investigators also found that the AI process used for the political messaging had been applied to the promotion of Kenyan retail businesses. The marketing activity was carried out using the personas “SHANKI” and “Elkins Marketer”.
The findings described the operation as an entirely domestic Kenyan campaign. Investigators noted that its pro-administration messaging and use of specific hashtags could indicate an alignment with the ruling coalition, but stopped short of identifying the group or organisation responsible.
According to the report, the campaign mixed messages supporting the government with attacks against the opposition, creating a pattern that investigators said was consistent with organised electoral communication.
The political content also drew from a system originally used for retail marketing. Promotional messages were adapted and presented as though they were independent comments from social media users rather than planned campaign material.
Investigators said Claude was regularly used to “humanise” and refine batches containing 50 pre-written topics and tweets.
The report said the AI tool mainly helped the actor produce content at a higher volume and make the posts look more natural. The main political themes, however, had already been settled before the content was processed through the chatbot.
The operation was classified as Category One on the Breakout Scale. This meant the activity was limited to a network of fake accounts and local influences on one platform, with investigators finding that it did not succeed in reaching or influencing real people.
The accounts and organisation connected to the operation were removed after an internal investigation. The probe was launched following a tip shared by OpenAI over repeat activity on its platform.
The investigation also helped investigators develop detection methods based on the behaviour and patterns observed during the operation.
The report comes as the government continues to face public attention over electricity tariffs and its decisions concerning proposed changes to power prices.